File S9 - Final Pareto Fronts for interactions 0-9 for all replicate runs for Experiment 3: Random features to XOR interactions ************************************************************************************** Random Seed 24 - 0 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00266 Best score on D1-D2 diff: 3.53124 Gen 2 - Best score on D2: -0.00005 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: -0.00005 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.00099 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00099 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00099 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00099 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.00099 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.00099 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.00099 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.00191 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.00191 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.00191 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.00191 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.00191 Best score on D1-D2 diff: 13.56244 Gen 16 - Best score on D2: 0.00191 Best score on D1-D2 diff: 13.56244 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.012553046048354277 Entire dataset(80%) R^2 trained on data (80%): 0.005474425142629857 Holdout R^2 (20%) trained on data (80%): -0.0037466774521557333 Dataset D1 R^2 on trained D1: 0.013339240089883964 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004389573351336384 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.001907195805677775 | D1-D2 diff: 6.285107105348616 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), SelectPercentile(percentile=45), HeterosisEncoder(), VarianceThreshold(threshold=0.2), RandomForestRegressor(max_features=0.55, min_samples_leaf=20, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002155160347820817 Holdout data R^2 trained on entire dataset(80%): -0.002072863177345674 Dataset D1 R^2 on trained D1: 0.0012663563126444899 .................................................. Pipeline #2: Score on D2: 0.000988104885820018 | D1-D2 diff: 9.778454659128222 Pipeline steps: VarianceThreshold(threshold=0.05), DominantEncoder(), SelectPercentile(percentile=50), VarianceThreshold(threshold=0.2), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.001067717403290036 Holdout data R^2 trained on entire dataset(80%): -0.0012851907934343654 Dataset D1 R^2 on trained D1: 0.0010974801448526694 .................................................. Pipeline #3: Score on D2: 0.00021645841093220763 | D1-D2 diff: 10.050295305573249 Pipeline steps: SelectPercentile(percentile=10), OverDominanceEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00013498438567838278 Holdout data R^2 trained on entire dataset(80%): 0.0002273342713182691 Dataset D1 R^2 on trained D1: 0.0003144716426528582 .................................................. Pipeline #4: Score on D2: -3.2015594250811574e-05 | D1-D2 diff: 13.480151627526027 Pipeline steps: RecessiveEncoder(), DominantEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=12, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021467281316e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516763899763 Dataset D1 R^2 on trained D1: -1.73106056000627e-06 .................................................. Pipeline #5: Score on D2: -3.307746938241429e-05 | D1-D2 diff: 13.562441741109108 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), SelectPercentile(percentile=65), SelectPercentile(percentile=45), HeterosisEncoder(), VarianceThreshold(threshold=0.2), RandomForestRegressor(max_features=0.55, min_samples_leaf=4, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0018373322255911262 Holdout data R^2 trained on entire dataset(80%): -0.0028345546172570124 Dataset D1 R^2 on trained D1: -3.521278857387955e-06 .................................................. ************************************************************************************** Random Seed 24 - 1 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.03783 Best score on D1-D2 diff: 3.53124 Gen 2 - Best score on D2: 0.03798 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.05211 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.05750 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.05750 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.05962 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.05962 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.05962 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.05962 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.05962 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.07428 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.07684 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.07684 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.07684 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.07687 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.08034 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.08034 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.08034 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.08034 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.08034 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.08034 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.08034 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.08034 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.08034 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.08034 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.014383662880555814 Entire dataset(80%) R^2 trained on data (80%): 0.005131604881063789 Holdout R^2 (20%) trained on data (80%): -0.0017809775060435573 Dataset D1 R^2 on trained D1: 0.013687437821544357 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004380065471687122 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.0803385303777604 | D1-D2 diff: 2.019770539280286 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=16, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0289538917331148 Holdout data R^2 trained on entire dataset(80%): -0.006988833111341686 Dataset D1 R^2 on trained D1: 0.14042710016000126 .................................................. Pipeline #2: Score on D2: 0.07547185857767413 | D1-D2 diff: 2.0312027370964016 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.028412080202453338 Holdout data R^2 trained on entire dataset(80%): -0.005175052699275318 Dataset D1 R^2 on trained D1: 0.13421902279660214 .................................................. Pipeline #3: Score on D2: 0.07376622393516163 | D1-D2 diff: 2.0384565028370565 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=16, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03196240626781144 Holdout data R^2 trained on entire dataset(80%): -0.004470866365507531 Dataset D1 R^2 on trained D1: 0.13168164322465037 .................................................. Pipeline #4: Score on D2: 0.07133378314885808 | D1-D2 diff: 2.091831744612518 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.028216035850369692 Holdout data R^2 trained on entire dataset(80%): -0.0025724420295598627 Dataset D1 R^2 on trained D1: 0.12356053332613004 .................................................. Pipeline #5: Score on D2: 0.0655778172443382 | D1-D2 diff: 2.092147869587713 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.027979101569298392 Holdout data R^2 trained on entire dataset(80%): -0.0021875230230428144 Dataset D1 R^2 on trained D1: 0.1177730085845472 .................................................. Pipeline #6: Score on D2: 0.051903129896402556 | D1-D2 diff: 2.1279066494450625 Pipeline steps: SelectPercentile(percentile=90), SelectPercentile(percentile=70), HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=14, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03791768126998796 Holdout data R^2 trained on entire dataset(80%): -0.007621790652038607 Dataset D1 R^2 on trained D1: 0.10067728031022938 .................................................. Pipeline #7: Score on D2: 0.04901885658453686 | D1-D2 diff: 2.159118171850817 Pipeline steps: SelectPercentile(percentile=55), HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03225689012811184 Holdout data R^2 trained on entire dataset(80%): -0.00748965660815637 Dataset D1 R^2 on trained D1: 0.09503331869216969 .................................................. Pipeline #8: Score on D2: 0.048784739291166224 | D1-D2 diff: 2.1722235349894685 Pipeline steps: SelectPercentile(percentile=55), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.031707153362223184 Holdout data R^2 trained on entire dataset(80%): -0.005056541030988626 Dataset D1 R^2 on trained D1: 0.09369876062864424 .................................................. Pipeline #9: Score on D2: 0.021787272215293085 | D1-D2 diff: 2.2489343017541543 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0660852569044692 Holdout data R^2 trained on entire dataset(80%): 0.025721986422411858 Dataset D1 R^2 on trained D1: 0.06087972549426923 .................................................. Pipeline #10: Score on D2: 0.018223805698016293 | D1-D2 diff: 2.2575477030504616 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), VarianceThreshold(threshold=0.35), RandomForestRegressor(max_features=0.05, min_samples_leaf=19, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.062194318547641236 Holdout data R^2 trained on entire dataset(80%): 0.021149242267343937 Dataset D1 R^2 on trained D1: 0.05672305455755999 .................................................. Pipeline #11: Score on D2: 0.0018911867607409594 | D1-D2 diff: 3.2676978373721606 Pipeline steps: SelectPercentile(percentile=90), SelectPercentile(percentile=70), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=14, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.008649390428918435 Holdout data R^2 trained on entire dataset(80%): 0.00048046077863328485 Dataset D1 R^2 on trained D1: 0.010661850024120567 .................................................. Pipeline #12: Score on D2: 0.0014261918607164636 | D1-D2 diff: 3.6567818559488123 Pipeline steps: SelectPercentile(percentile=90), SelectPercentile(percentile=60), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=14, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.009446342815498832 Holdout data R^2 trained on entire dataset(80%): -0.0002558804108681034 Dataset D1 R^2 on trained D1: 0.0070186634476688825 .................................................. Pipeline #13: Score on D2: 0.0006175093762322836 | D1-D2 diff: 5.332206896208156 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), OverDominanceEncoder(), SelectPercentile(percentile=35), FeatureEncodingFrequencySelector(threshold=0.25), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=16, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0030763567539595638 Holdout data R^2 trained on entire dataset(80%): -0.005803219304407969 Dataset D1 R^2 on trained D1: 0.0018545160169093355 .................................................. Pipeline #14: Score on D2: 0.00021645841644202246 | D1-D2 diff: 10.05029530731504 Pipeline steps: SelectPercentile(percentile=10), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0010264239288023669 Holdout data R^2 trained on entire dataset(80%): -0.002396959021580436 Dataset D1 R^2 on trained D1: 0.0003144716480947274 .................................................. Pipeline #15: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=11, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 24 - 2 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.09217 Best score on D1-D2 diff: 4.28569 Gen 2 - Best score on D2: 0.09745 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.09745 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.09745 Best score on D1-D2 diff: 11.96036 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.011399938128363152 Entire dataset(80%) R^2 trained on data (80%): 0.004606413487589389 Holdout R^2 (20%) trained on data (80%): -0.0027810666206373735 Dataset D1 R^2 on trained D1: 0.010775467817440987 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003807371439866314 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.09744767747094663 | D1-D2 diff: 1.6220180150731065 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=13, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24548893638000968 Holdout data R^2 trained on entire dataset(80%): 0.11691675303732518 Dataset D1 R^2 on trained D1: 0.24191755555285255 .................................................. Pipeline #2: Score on D2: 0.09288391946003216 | D1-D2 diff: 1.7210069098444638 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21350357775640072 Holdout data R^2 trained on entire dataset(80%): 0.11458387010320459 Dataset D1 R^2 on trained D1: 0.20687465182301146 .................................................. Pipeline #3: Score on D2: 0.08656564049366755 | D1-D2 diff: 2.2700033450208825 Pipeline steps: DecisionTreeRegressor(max_depth=4, min_samples_leaf=6, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.016146000388248205 Holdout data R^2 trained on entire dataset(80%): -0.0030317296915041148 Dataset D1 R^2 on trained D1: 0.12422682813835617 .................................................. Pipeline #4: Score on D2: 0.06724234199563739 | D1-D2 diff: 2.3727598897832363 Pipeline steps: DecisionTreeRegressor(max_depth=3, min_samples_leaf=18, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.007033391338790107 Holdout data R^2 trained on entire dataset(80%): -0.012530121406759465 Dataset D1 R^2 on trained D1: 0.09879128627596878 .................................................. Pipeline #5: Score on D2: 0.019345583652769283 | D1-D2 diff: 3.5258609983823237 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0032153718811361864 Holdout data R^2 trained on entire dataset(80%): -0.0009604869077197709 Dataset D1 R^2 on trained D1: 0.025816104672262363 .................................................. Pipeline #6: Score on D2: -4.886780840163141e-05 | D1-D2 diff: 11.960360895673112 Pipeline steps: RecessiveEncoder(), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0 Holdout data R^2 trained on entire dataset(80%): -0.0012028184006112053 Dataset D1 R^2 on trained D1: 0.0 .................................................. ************************************************************************************** Random Seed 24 - 3 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.11449 Best score on D1-D2 diff: 6.23901 Gen 2 - Best score on D2: 0.11894 Best score on D1-D2 diff: 6.23901 Gen 3 - Best score on D2: 0.12106 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.12106 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.12422 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.12422 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.12422 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.12422 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.12422 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.12422 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.12426 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.12426 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.12426 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.12426 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.12426 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.12426 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.12426 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.12649 Best score on D1-D2 diff: 24.43121 Gen 19 - Best score on D2: 0.12649 Best score on D1-D2 diff: 24.43121 Gen 20 - Best score on D2: 0.12649 Best score on D1-D2 diff: 24.43121 Gen 21 - Best score on D2: 0.12649 Best score on D1-D2 diff: 24.43121 Gen 22 - Best score on D2: 0.12723 Best score on D1-D2 diff: 24.43121 Gen 23 - Best score on D2: 0.12723 Best score on D1-D2 diff: 24.43121 Gen 24 - Best score on D2: 0.12723 Best score on D1-D2 diff: 24.43121 Gen 25 - Best score on D2: 0.13220 Best score on D1-D2 diff: 24.43121 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.006485665064594803 Entire dataset(80%) R^2 trained on data (80%): 0.006125266856127798 Holdout R^2 (20%) trained on data (80%): 0.001792620662578459 Dataset D1 R^2 on trained D1: 0.00981364041053967 Combined Dataset (100%) R^2 trained on combined data (100%): 0.006164087606294832 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.13219982370163363 | D1-D2 diff: 1.9601816810543242 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), RecessiveEncoder(), HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.9000000000000001, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21737657612750505 Holdout data R^2 trained on entire dataset(80%): 0.14502480169151588 Dataset D1 R^2 on trained D1: 0.19993506704014596 .................................................. Pipeline #2: Score on D2: 0.1152534730422512 | D1-D2 diff: 1.9851767843311958 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RecessiveEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=18, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20015759044865644 Holdout data R^2 trained on entire dataset(80%): 0.13860340990317654 Dataset D1 R^2 on trained D1: 0.1796412230549177 .................................................. Pipeline #3: Score on D2: 0.11508571223952913 | D1-D2 diff: 2.1698429847253546 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), SelectPercentile(percentile=85), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1770711014179065 Holdout data R^2 trained on entire dataset(80%): 0.12746022592866946 Dataset D1 R^2 on trained D1: 0.16019716015400154 .................................................. Pipeline #4: Score on D2: 0.10262931327610714 | D1-D2 diff: 2.9800363982366678 Pipeline steps: VarianceThreshold(threshold=0.3), VarianceThreshold(threshold=0.2), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=3, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11651860284625826 Holdout data R^2 trained on entire dataset(80%): 0.09339940056220386 Dataset D1 R^2 on trained D1: 0.11530915187986412 .................................................. Pipeline #5: Score on D2: 0.09289592315102424 | D1-D2 diff: 4.427247338032546 Pipeline steps: SelectPercentile(percentile=70), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=9, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.008341518101785295 Holdout data R^2 trained on entire dataset(80%): -0.007328591787236771 Dataset D1 R^2 on trained D1: 0.09029297895140909 .................................................. Pipeline #6: Score on D2: 0.09289592315102413 | D1-D2 diff: 4.427247338032593 Pipeline steps: SelectPercentile(percentile=85), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=9, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10653657331495414 Holdout data R^2 trained on entire dataset(80%): 0.06964587195231087 Dataset D1 R^2 on trained D1: 0.09029297895140909 .................................................. Pipeline #7: Score on D2: 0.0919253392125926 | D1-D2 diff: 4.790541140066924 Pipeline steps: SelectPercentile(percentile=70), OverDominanceEncoder(), VarianceThreshold(threshold=0.25), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.008341518101785295 Holdout data R^2 trained on entire dataset(80%): -0.007328591787236771 Dataset D1 R^2 on trained D1: 0.090026615809355 .................................................. Pipeline #8: Score on D2: 0.0919253392125925 | D1-D2 diff: 4.790541140066995 Pipeline steps: SelectPercentile(percentile=70), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.008341518101785295 Holdout data R^2 trained on entire dataset(80%): -0.007328591787236771 Dataset D1 R^2 on trained D1: 0.090026615809355 .................................................. Pipeline #9: Score on D2: 0.09141147512183079 | D1-D2 diff: 5.395545350739225 Pipeline steps: SelectPercentile(percentile=70), UnderDominanceEncoder(), VarianceThreshold(threshold=0.3), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01689944199238491 Holdout data R^2 trained on entire dataset(80%): 0.005280825772159936 Dataset D1 R^2 on trained D1: 0.09023153868884082 .................................................. Pipeline #10: Score on D2: 0.09141147512183068 | D1-D2 diff: 5.395545350739352 Pipeline steps: SelectPercentile(percentile=70), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=18, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01689944199238491 Holdout data R^2 trained on entire dataset(80%): 0.005280825772159825 Dataset D1 R^2 on trained D1: 0.09023153868884082 .................................................. Pipeline #11: Score on D2: 0.08976702390612112 | D1-D2 diff: 6.239011670416305 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10542309363446589 Holdout data R^2 trained on entire dataset(80%): 0.07056225020369533 Dataset D1 R^2 on trained D1: 0.08910703473781956 .................................................. Pipeline #12: Score on D2: 0.08909240618715697 | D1-D2 diff: 6.4215260460001184 Pipeline steps: SelectPercentile(percentile=70), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.007783502876056114 Holdout data R^2 trained on entire dataset(80%): -0.006715530228864397 Dataset D1 R^2 on trained D1: 0.08850431183806473 .................................................. Pipeline #13: Score on D2: 0.08835609095234798 | D1-D2 diff: 6.57034144054427 Pipeline steps: SelectPercentile(percentile=55), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.007748620864615252 Holdout data R^2 trained on entire dataset(80%): -0.002617509733229495 Dataset D1 R^2 on trained D1: 0.08781949394827271 .................................................. Pipeline #14: Score on D2: 0.0014136852328071603 | D1-D2 diff: 9.382792741726059 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), VarianceThreshold(threshold=0.35), SelectPercentile(percentile=35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.001373337617810222 Holdout data R^2 trained on entire dataset(80%): -0.00034102929191326403 Dataset D1 R^2 on trained D1: 0.001284660934123938 .................................................. Pipeline #15: Score on D2: 0.0013062130014446227 | D1-D2 diff: 24.431212821310144 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), VarianceThreshold(threshold=0.35), SelectPercentile(percentile=85), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.001375114814078704 Holdout data R^2 trained on entire dataset(80%): -0.00043562737468105794 Dataset D1 R^2 on trained D1: 0.0013090198558926014 .................................................. ************************************************************************************** Random Seed 24 - 4 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.14585 Best score on D1-D2 diff: 5.47056 Gen 2 - Best score on D2: 0.15440 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.15562 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15562 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.15562 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.15562 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15562 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.15562 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.15562 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.15564 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.15620 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.16336 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.16368 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.16368 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.16368 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.01583710878815925 Entire dataset(80%) R^2 trained on data (80%): 0.0041263065066986515 Holdout R^2 (20%) trained on data (80%): -0.002289364620909451 Dataset D1 R^2 on trained D1: 0.01386453152495104 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004027278524161737 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.16368295895417784 | D1-D2 diff: 1.9199515385001544 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), SelectPercentile(percentile=65), FeatureEncodingFrequencySelector(threshold=0.3), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=19, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21815543960943085 Holdout data R^2 trained on entire dataset(80%): 0.0999260864722975 Dataset D1 R^2 on trained D1: 0.23727636997513768 .................................................. Pipeline #2: Score on D2: 0.15841462879018653 | D1-D2 diff: 1.9714259730477126 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), SelectPercentile(percentile=75), FeatureEncodingFrequencySelector(threshold=0.3), RandomForestRegressor(max_features=0.3, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22706964375293304 Holdout data R^2 trained on entire dataset(80%): 0.11861066866594461 Dataset D1 R^2 on trained D1: 0.22461769495976203 .................................................. Pipeline #3: Score on D2: 0.14430345004018097 | D1-D2 diff: 1.9787251676235578 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), SelectPercentile(percentile=55), FeatureEncodingFrequencySelector(threshold=0.3), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19782826681178367 Holdout data R^2 trained on entire dataset(80%): 0.10651083059925304 Dataset D1 R^2 on trained D1: 0.20953505881084666 .................................................. Pipeline #4: Score on D2: 0.13408728396490066 | D1-D2 diff: 1.995178913492256 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=35), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=19, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17221848340191537 Holdout data R^2 trained on entire dataset(80%): 0.09290465698981054 Dataset D1 R^2 on trained D1: 0.19719356906048724 .................................................. Pipeline #5: Score on D2: 0.12249800596226723 | D1-D2 diff: 2.1299850158622666 Pipeline steps: VarianceThreshold(threshold=0.35), SelectPercentile(percentile=65), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21228291887447326 Holdout data R^2 trained on entire dataset(80%): 0.09942830783374257 Dataset D1 R^2 on trained D1: 0.17108206624660938 .................................................. Pipeline #6: Score on D2: 0.10394922363516701 | D1-D2 diff: 3.229173991528639 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10871053168981604 Holdout data R^2 trained on entire dataset(80%): 0.060752507274451584 Dataset D1 R^2 on trained D1: 0.11314597014771854 .................................................. Pipeline #7: Score on D2: 0.054686182277618456 | D1-D2 diff: 4.263238162913153 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011033232849264829 Holdout data R^2 trained on entire dataset(80%): -0.007818806663438593 Dataset D1 R^2 on trained D1: 0.057713385802616646 .................................................. Pipeline #8: Score on D2: 0.007149890511288004 | D1-D2 diff: 5.635617667818145 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.006784721975439778 Holdout data R^2 trained on entire dataset(80%): -0.003195383874071922 Dataset D1 R^2 on trained D1: 0.006158524758154482 .................................................. Pipeline #9: Score on D2: 0.0008104898731055021 | D1-D2 diff: 5.78851060757743 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), DecisionTreeRegressor(max_depth=1, min_samples_leaf=9, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.001318293446541885 Holdout data R^2 trained on entire dataset(80%): -0.0015222547954638621 Dataset D1 R^2 on trained D1: 0.0017011923624142833 .................................................. Pipeline #10: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), DominantEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 24 - 5 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.13883 Best score on D1-D2 diff: 4.22481 Gen 2 - Best score on D2: 0.14143 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14531 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.14879 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.14879 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.14879 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.14879 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.14879 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.14879 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.14879 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.14879 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.14879 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.15084 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.15084 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.15084 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.15084 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.15084 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.15158 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.15158 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.15158 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.15158 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.15158 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.15158 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.15158 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.15158 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.01522197604931863 Entire dataset(80%) R^2 trained on data (80%): 0.003563186785491168 Holdout R^2 (20%) trained on data (80%): -0.004459922055877419 Dataset D1 R^2 on trained D1: 0.010403695244665556 Combined Dataset (100%) R^2 trained on combined data (100%): 0.002903490896122629 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.15158236741275266 | D1-D2 diff: 1.4727605594442703 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=2, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.36985943295531076 Holdout data R^2 trained on entire dataset(80%): 0.17095694781692694 Dataset D1 R^2 on trained D1: 0.3641374132123987 .................................................. Pipeline #2: Score on D2: 0.1508332125577797 | D1-D2 diff: 1.4838477252018483 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), OverDominanceEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=2, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3712735078734726 Holdout data R^2 trained on entire dataset(80%): 0.17395219971800346 Dataset D1 R^2 on trained D1: 0.357106342739144 .................................................. Pipeline #3: Score on D2: 0.1488457513431949 | D1-D2 diff: 1.5227518028648495 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=2, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33929384281635877 Holdout data R^2 trained on entire dataset(80%): 0.1697968151295317 Dataset D1 R^2 on trained D1: 0.3348331459032279 .................................................. Pipeline #4: Score on D2: 0.1466011733355661 | D1-D2 diff: 1.5603231076689352 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=11, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3170270931204171 Holdout data R^2 trained on entire dataset(80%): 0.1627585117664323 Dataset D1 R^2 on trained D1: 0.3153115670886093 .................................................. Pipeline #5: Score on D2: 0.1463721132026118 | D1-D2 diff: 1.5760157573450682 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=3, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3174089433772824 Holdout data R^2 trained on entire dataset(80%): 0.16848085500466947 Dataset D1 R^2 on trained D1: 0.30846269407807114 .................................................. Pipeline #6: Score on D2: 0.1443151149364943 | D1-D2 diff: 1.7075566687893124 Pipeline steps: SelectPercentile(percentile=90), VarianceThreshold(threshold=0.15), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=13, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26975566474119694 Holdout data R^2 trained on entire dataset(80%): 0.17038554345638468 Dataset D1 R^2 on trained D1: 0.26194007738753045 .................................................. Pipeline #7: Score on D2: 0.14172591570995996 | D1-D2 diff: 1.7417217246042822 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=14, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25785581855132855 Holdout data R^2 trained on entire dataset(80%): 0.16604141648997228 Dataset D1 R^2 on trained D1: 0.25038972477417587 .................................................. Pipeline #8: Score on D2: 0.13856463018915066 | D1-D2 diff: 1.7540230259543852 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=16, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2555709220460991 Holdout data R^2 trained on entire dataset(80%): 0.16373370424402212 Dataset D1 R^2 on trained D1: 0.2442120363799939 .................................................. Pipeline #9: Score on D2: 0.13768238375791597 | D1-D2 diff: 1.7661707380771514 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=17, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24866181940995202 Holdout data R^2 trained on entire dataset(80%): 0.16390436997826363 Dataset D1 R^2 on trained D1: 0.240453070736516 .................................................. Pipeline #10: Score on D2: 0.13746534824823076 | D1-D2 diff: 1.7713530252576564 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=17, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25003483648650937 Holdout data R^2 trained on entire dataset(80%): 0.16621234489573566 Dataset D1 R^2 on trained D1: 0.23903863516486723 .................................................. Pipeline #11: Score on D2: 0.13743050244353716 | D1-D2 diff: 1.7785781197937547 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), OverDominanceEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24601618359694655 Holdout data R^2 trained on entire dataset(80%): 0.16814316476571833 Dataset D1 R^2 on trained D1: 0.2373633399145274 .................................................. Pipeline #12: Score on D2: 0.13720360094276274 | D1-D2 diff: 1.7844101685438565 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24636065440853627 Holdout data R^2 trained on entire dataset(80%): 0.16653109501294572 Dataset D1 R^2 on trained D1: 0.23583637373521216 .................................................. Pipeline #13: Score on D2: 0.1367166629678338 | D1-D2 diff: 1.8431443562529266 Pipeline steps: VarianceThreshold(threshold=0.05), HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23801275179383496 Holdout data R^2 trained on entire dataset(80%): 0.1648860779574911 Dataset D1 R^2 on trained D1: 0.2233654718179653 .................................................. Pipeline #14: Score on D2: 0.13596021969097594 | D1-D2 diff: 1.9230536131244815 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22081182734070548 Holdout data R^2 trained on entire dataset(80%): 0.16021580746053843 Dataset D1 R^2 on trained D1: 0.20907992481356452 .................................................. Pipeline #15: Score on D2: 0.12152463964130888 | D1-D2 diff: 2.5923971178252176 Pipeline steps: SelectPercentile(percentile=50), VarianceThreshold(threshold=0.3), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=17, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14095462184925378 Holdout data R^2 trained on entire dataset(80%): 0.1412914577494091 Dataset D1 R^2 on trained D1: 0.1436654687673482 .................................................. Pipeline #16: Score on D2: 0.09969672833773846 | D1-D2 diff: 4.28650750060088 Pipeline steps: SelectPercentile(percentile=50), VarianceThreshold(threshold=0.25), DecisionTreeRegressor(max_depth=4, min_samples_leaf=20, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.031012960282010615 Holdout data R^2 trained on entire dataset(80%): 0.002391418667621159 Dataset D1 R^2 on trained D1: 0.10265873238773815 .................................................. Pipeline #17: Score on D2: 0.08472370237342597 | D1-D2 diff: 6.211748480472145 Pipeline steps: SelectPercentile(percentile=75), HeterosisEncoder(), SelectPercentile(percentile=95), DecisionTreeRegressor(max_depth=2, min_samples_leaf=17, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08682026836452694 Holdout data R^2 trained on entire dataset(80%): 0.12466595702661432 Dataset D1 R^2 on trained D1: 0.08405205000532767 .................................................. Pipeline #18: Score on D2: 0.08472370237342586 | D1-D2 diff: 6.2117484804724015 Pipeline steps: SelectPercentile(percentile=50), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=20, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08868483134939131 Holdout data R^2 trained on entire dataset(80%): 0.10847292151805776 Dataset D1 R^2 on trained D1: 0.08405205000532767 .................................................. Pipeline #19: Score on D2: 0.08472370237342575 | D1-D2 diff: 6.211748480472658 Pipeline steps: SelectPercentile(percentile=75), HeterosisEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=17, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08682026836452694 Holdout data R^2 trained on entire dataset(80%): 0.12466595702661443 Dataset D1 R^2 on trained D1: 0.08405205000532767 .................................................. Pipeline #20: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: RecessiveEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.7500000000000001, min_samples_leaf=8, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 24 - 6 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.16852 Best score on D1-D2 diff: 5.41450 Gen 2 - Best score on D2: 0.17062 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.17062 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.17062 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.17062 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.17062 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.17062 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.17062 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.17062 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.17373 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.17373 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.17373 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.17373 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.17508 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.17508 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.17508 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.17508 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.17508 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.17550 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.17550 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.17550 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.17550 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.17550 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.17550 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.17550 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.009545631589012338 Entire dataset(80%) R^2 trained on data (80%): 0.0064507718766292355 Holdout R^2 (20%) trained on data (80%): -0.0012382374277555286 Dataset D1 R^2 on trained D1: 0.011323952262901171 Combined Dataset (100%) R^2 trained on combined data (100%): 0.00560354702085708 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.17550175996773698 | D1-D2 diff: 1.4076624611421487 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.43279662210269143 Holdout data R^2 trained on entire dataset(80%): 0.1913187067170179 Dataset D1 R^2 on trained D1: 0.4301882353039188 .................................................. Pipeline #2: Score on D2: 0.17521213281280557 | D1-D2 diff: 1.4760393556617557 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.39121399953801816 Holdout data R^2 trained on entire dataset(80%): 0.19272110439336487 Dataset D1 R^2 on trained D1: 0.3858848278560628 .................................................. Pipeline #3: Score on D2: 0.17508138976006782 | D1-D2 diff: 1.5660362692866954 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3481418123246103 Holdout data R^2 trained on entire dataset(80%): 0.19125719169110011 Dataset D1 R^2 on trained D1: 0.3413432884304092 .................................................. Pipeline #4: Score on D2: 0.1737258171832715 | D1-D2 diff: 1.6359234036777526 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3215870851735846 Holdout data R^2 trained on entire dataset(80%): 0.1841627491558252 Dataset D1 R^2 on trained D1: 0.31334597930639496 .................................................. Pipeline #5: Score on D2: 0.17095054075669047 | D1-D2 diff: 1.6504841853937546 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), FeatureEncodingFrequencySelector(threshold=0.0), RecessiveEncoder(), VarianceThreshold(), VarianceThreshold(threshold=0.05), RandomForestRegressor(max_features=0.25, min_samples_leaf=6, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3182086781874264 Holdout data R^2 trained on entire dataset(80%): 0.18655741716390373 Dataset D1 R^2 on trained D1: 0.305708532606184 .................................................. Pipeline #6: Score on D2: 0.17046220607536233 | D1-D2 diff: 1.8009257448725493 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2736134808459082 Holdout data R^2 trained on entire dataset(80%): 0.1778118882656231 Dataset D1 R^2 on trained D1: 0.2655263570743114 .................................................. Pipeline #7: Score on D2: 0.1696165500921648 | D1-D2 diff: 1.8520681281440237 Pipeline steps: SelectPercentile(percentile=95), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26171026670376873 Holdout data R^2 trained on entire dataset(80%): 0.18177791429330703 Dataset D1 R^2 on trained D1: 0.25460739908382957 .................................................. Pipeline #8: Score on D2: 0.16426160833984682 | D1-D2 diff: 1.8696568898739083 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=17, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25611406164836403 Holdout data R^2 trained on entire dataset(80%): 0.1731053705672716 Dataset D1 R^2 on trained D1: 0.24609910625228681 .................................................. Pipeline #9: Score on D2: 0.1633970658748397 | D1-D2 diff: 1.9133741929247683 Pipeline steps: SelectPercentile(percentile=90), HeterosisEncoder(), RecessiveEncoder(), VarianceThreshold(threshold=0.05), SelectPercentile(percentile=95), RandomForestRegressor(max_features=0.2, min_samples_leaf=10, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2503759182757419 Holdout data R^2 trained on entire dataset(80%): 0.16463080170059752 Dataset D1 R^2 on trained D1: 0.23800763485351706 .................................................. Pipeline #10: Score on D2: 0.16183115029517758 | D1-D2 diff: 1.9214534566948402 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=14, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24678910033629664 Holdout data R^2 trained on entire dataset(80%): 0.16908731732543325 Dataset D1 R^2 on trained D1: 0.23519473164469973 .................................................. Pipeline #11: Score on D2: 0.1612784273896315 | D1-D2 diff: 1.9749798618413756 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=16, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2399226592295317 Holdout data R^2 trained on entire dataset(80%): 0.1705417462049127 Dataset D1 R^2 on trained D1: 0.22700626028400817 .................................................. Pipeline #12: Score on D2: 0.15936859907074064 | D1-D2 diff: 1.9917364640290693 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=16, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23446797729273994 Holdout data R^2 trained on entire dataset(80%): 0.16762698266280263 Dataset D1 R^2 on trained D1: 0.2229122995699243 .................................................. Pipeline #13: Score on D2: 0.15935884192115957 | D1-D2 diff: 2.022916076741683 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), VarianceThreshold(threshold=0.05), RandomForestRegressor(max_features=0.25, min_samples_leaf=16, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24053213807875473 Holdout data R^2 trained on entire dataset(80%): 0.17530302041243107 Dataset D1 R^2 on trained D1: 0.21907454314588415 .................................................. Pipeline #14: Score on D2: 0.15902679185626023 | D1-D2 diff: 2.027250956262954 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=85), RandomForestRegressor(max_features=0.4, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2392602334931644 Holdout data R^2 trained on entire dataset(80%): 0.16799907430161398 Dataset D1 R^2 on trained D1: 0.21823336761898726 .................................................. Pipeline #15: Score on D2: 0.15669752090208855 | D1-D2 diff: 2.0476001319213695 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22749293835421625 Holdout data R^2 trained on entire dataset(80%): 0.16371747866644049 Dataset D1 R^2 on trained D1: 0.2135853557158751 .................................................. Pipeline #16: Score on D2: 0.15278100003350514 | D1-D2 diff: 2.2361547536726207 Pipeline steps: SelectPercentile(percentile=90), HeterosisEncoder(), RecessiveEncoder(), VarianceThreshold(threshold=0.05), SelectPercentile(percentile=95), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.2, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21320711094198552 Holdout data R^2 trained on entire dataset(80%): 0.1573180202931469 Dataset D1 R^2 on trained D1: 0.19277479143838583 .................................................. Pipeline #17: Score on D2: 0.14113492248002746 | D1-D2 diff: 2.317261390812054 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=60), RandomForestRegressor(max_features=0.4, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21546895401003485 Holdout data R^2 trained on entire dataset(80%): 0.1635032201865123 Dataset D1 R^2 on trained D1: 0.1758165841560192 .................................................. Pipeline #18: Score on D2: 0.14086452128908378 | D1-D2 diff: 2.388497316633719 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), SelectPercentile(percentile=60), RandomForestRegressor(max_features=0.25, min_samples_leaf=14, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1834485939833308 Holdout data R^2 trained on entire dataset(80%): 0.13452889768990128 Dataset D1 R^2 on trained D1: 0.17159016350824063 .................................................. Pipeline #19: Score on D2: 0.12849765900138066 | D1-D2 diff: 2.4101916034957203 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=60), RandomForestRegressor(max_features=0.25, min_samples_leaf=14, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20747287832376204 Holdout data R^2 trained on entire dataset(80%): 0.15847172336220172 Dataset D1 R^2 on trained D1: 0.1581318941992378 .................................................. Pipeline #20: Score on D2: 0.1257290135211655 | D1-D2 diff: 3.190415495941412 Pipeline steps: SelectPercentile(percentile=30), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1553165082054575 Holdout data R^2 trained on entire dataset(80%): 0.11518080428005972 Dataset D1 R^2 on trained D1: 0.1353808735673463 .................................................. Pipeline #21: Score on D2: 0.11631397770039087 | D1-D2 diff: 3.2945500710329396 Pipeline steps: SelectPercentile(percentile=30), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12722890141147447 Holdout data R^2 trained on entire dataset(80%): 0.09926213503411141 Dataset D1 R^2 on trained D1: 0.10782577874645327 .................................................. Pipeline #22: Score on D2: 0.1156732294171906 | D1-D2 diff: 4.092178288919032 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=70), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14689164117669173 Holdout data R^2 trained on entire dataset(80%): 0.12531423464695257 Dataset D1 R^2 on trained D1: 0.11923923330595854 .................................................. Pipeline #23: Score on D2: 0.07161311582687113 | D1-D2 diff: 4.952162646588748 Pipeline steps: VarianceThreshold(threshold=0.2), SelectPercentile(percentile=70), DecisionTreeRegressor(max_depth=4, min_samples_leaf=8, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08306528822644055 Holdout data R^2 trained on entire dataset(80%): 0.07372109915654779 Dataset D1 R^2 on trained D1: 0.06995039092934574 .................................................. Pipeline #24: Score on D2: 0.0711372192144214 | D1-D2 diff: 7.112189873016598 Pipeline steps: VarianceThreshold(threshold=0.15), SelectPercentile(percentile=70), DecisionTreeRegressor(max_depth=4, min_samples_leaf=8, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1155667880394996 Holdout data R^2 trained on entire dataset(80%): 0.10990728406158612 Dataset D1 R^2 on trained D1: 0.07074639035037145 .................................................. Pipeline #25: Score on D2: 0.05963997832355783 | D1-D2 diff: 10.178330924740767 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.05), VarianceThreshold(threshold=0.1), SelectPercentile(percentile=30), RandomForestRegressor(max_features=0.5, min_samples_leaf=9, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1294933467174343 Holdout data R^2 trained on entire dataset(80%): 0.10183181399957442 Dataset D1 R^2 on trained D1: 0.05973315210736441 .................................................. Pipeline #26: Score on D2: 0.021761975276075818 | D1-D2 diff: 12.587570600368785 Pipeline steps: DominantEncoder(), VarianceThreshold(threshold=0.15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.025512818248347724 Holdout data R^2 trained on entire dataset(80%): 0.019775094002593252 Dataset D1 R^2 on trained D1: 0.021801807295252384 .................................................. Pipeline #27: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: OverDominanceEncoder(), RecessiveEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 24 - 7 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.19595 Best score on D1-D2 diff: 10.97805 Gen 2 - Best score on D2: 0.19763 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.20020 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.20448 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.20448 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.20448 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.20448 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.20448 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.20448 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.20448 Best score on D1-D2 diff: 21.14935 Gen 11 - Best score on D2: 0.20775 Best score on D1-D2 diff: 21.14935 Gen 12 - Best score on D2: 0.20775 Best score on D1-D2 diff: 21.14935 Gen 13 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 14 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 15 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 16 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 17 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 18 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 19 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 20 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 21 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 22 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 23 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 24 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 Gen 25 - Best score on D2: 0.20890 Best score on D1-D2 diff: 21.14935 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.009504722996563242 Entire dataset(80%) R^2 trained on data (80%): 0.004434250088211633 Holdout R^2 (20%) trained on data (80%): -0.004586083869223989 Dataset D1 R^2 on trained D1: 0.009607902453323214 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003372485525243163 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.20889569499353633 | D1-D2 diff: 1.5552699234448477 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=4, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.38768460119052406 Holdout data R^2 trained on entire dataset(80%): 0.19405821900442877 Dataset D1 R^2 on trained D1: 0.37980940680890507 .................................................. Pipeline #2: Score on D2: 0.20823831891860445 | D1-D2 diff: 1.6983486973920947 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34680723425529736 Holdout data R^2 trained on entire dataset(80%): 0.20128508830438563 Dataset D1 R^2 on trained D1: 0.32843502053345086 .................................................. Pipeline #3: Score on D2: 0.20257925219238204 | D1-D2 diff: 1.806611385330045 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(bootstrap=False, max_features=0.5, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30377691755538694 Holdout data R^2 trained on entire dataset(80%): 0.17231723194431392 Dataset D1 R^2 on trained D1: 0.29645232380751785 .................................................. Pipeline #4: Score on D2: 0.19945475599321905 | D1-D2 diff: 1.8071850748870908 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), RandomForestRegressor(max_features=0.55, min_samples_leaf=13, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3109635587328172 Holdout data R^2 trained on entire dataset(80%): 0.18834735141116798 Dataset D1 R^2 on trained D1: 0.29320868460926897 .................................................. Pipeline #5: Score on D2: 0.19944186683392007 | D1-D2 diff: 1.8106897547350946 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=13, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3114679806093533 Holdout data R^2 trained on entire dataset(80%): 0.19146200602506225 Dataset D1 R^2 on trained D1: 0.2924720386236749 .................................................. Pipeline #6: Score on D2: 0.1983988337441832 | D1-D2 diff: 1.8522634586440572 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=13, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29900924499065695 Holdout data R^2 trained on entire dataset(80%): 0.19086124484160305 Dataset D1 R^2 on trained D1: 0.28335383755604426 .................................................. Pipeline #7: Score on D2: 0.19774303799693793 | D1-D2 diff: 2.0459932256351108 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2801684929256001 Holdout data R^2 trained on entire dataset(80%): 0.18929000150422326 Dataset D1 R^2 on trained D1: 0.2548098004193786 .................................................. Pipeline #8: Score on D2: 0.19768354131842802 | D1-D2 diff: 2.0673705188637097 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.5, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2641398882501198 Holdout data R^2 trained on entire dataset(80%): 0.17190298288488326 Dataset D1 R^2 on trained D1: 0.25242630571394775 .................................................. Pipeline #9: Score on D2: 0.19392038873594175 | D1-D2 diff: 2.088006573495394 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26832248480899856 Holdout data R^2 trained on entire dataset(80%): 0.18551634162808706 Dataset D1 R^2 on trained D1: 0.2465309038048109 .................................................. Pipeline #10: Score on D2: 0.19351943641313107 | D1-D2 diff: 2.1140519200832344 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), VarianceThreshold(threshold=0.15), DominantEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=20, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26952756025364977 Holdout data R^2 trained on entire dataset(80%): 0.18500122864282742 Dataset D1 R^2 on trained D1: 0.2435848033296626 .................................................. Pipeline #11: Score on D2: 0.1924145472136224 | D1-D2 diff: 2.1868099930969493 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), SelectPercentile(percentile=85), RandomForestRegressor(max_features=0.4, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25128298596475795 Holdout data R^2 trained on entire dataset(80%): 0.1698797516532463 Dataset D1 R^2 on trained D1: 0.23614216322848192 .................................................. Pipeline #12: Score on D2: 0.19081014210820102 | D1-D2 diff: 2.240864343958581 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), HeterosisEncoder(), SelectPercentile(percentile=85), RandomForestRegressor(max_features=0.4, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24717675541062545 Holdout data R^2 trained on entire dataset(80%): 0.17262087531442327 Dataset D1 R^2 on trained D1: 0.2304687746065719 .................................................. Pipeline #13: Score on D2: 0.1825036425243215 | D1-D2 diff: 2.2598136807989726 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), RecessiveEncoder(), SelectPercentile(percentile=95), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23879337399819067 Holdout data R^2 trained on entire dataset(80%): 0.16932757515728147 Dataset D1 R^2 on trained D1: 0.2208487064347564 .................................................. Pipeline #14: Score on D2: 0.18017176782935262 | D1-D2 diff: 2.2856356178836457 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=80), RecessiveEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=19, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23841824650735566 Holdout data R^2 trained on entire dataset(80%): 0.15933089850182736 Dataset D1 R^2 on trained D1: 0.2168131644815774 .................................................. Pipeline #15: Score on D2: 0.17050361272840953 | D1-D2 diff: 2.299432836345814 Pipeline steps: VarianceThreshold(threshold=0.15), FeatureEncodingFrequencySelector(threshold=0.35), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22961082051025916 Holdout data R^2 trained on entire dataset(80%): 0.16956408939830936 Dataset D1 R^2 on trained D1: 0.20627345988994317 .................................................. Pipeline #16: Score on D2: 0.1673099065940793 | D1-D2 diff: 2.3581188978252445 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), HeterosisEncoder(), SelectPercentile(percentile=85), RandomForestRegressor(max_features=0.25, min_samples_leaf=19, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2222909606526886 Holdout data R^2 trained on entire dataset(80%): 0.16237924954172267 Dataset D1 R^2 on trained D1: 0.19964969731594118 .................................................. Pipeline #17: Score on D2: 0.16579078589584695 | D1-D2 diff: 2.3875223223164626 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=55), RandomForestRegressor(max_features=0.55, min_samples_leaf=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24566303970197578 Holdout data R^2 trained on entire dataset(80%): 0.1566746053738457 Dataset D1 R^2 on trained D1: 0.19656664868352958 .................................................. Pipeline #18: Score on D2: 0.16488095186269236 | D1-D2 diff: 2.3956507901885518 Pipeline steps: SelectPercentile(percentile=70), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=20, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24987445062389324 Holdout data R^2 trained on entire dataset(80%): 0.18507004161367013 Dataset D1 R^2 on trained D1: 0.19524124441092083 .................................................. Pipeline #19: Score on D2: 0.1643167492212544 | D1-D2 diff: 2.7380981658699635 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=55), RandomForestRegressor(max_features=0.5, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23198975262289045 Holdout data R^2 trained on entire dataset(80%): 0.15818354849219163 Dataset D1 R^2 on trained D1: 0.1821078959996606 .................................................. Pipeline #20: Score on D2: 0.15559227780949814 | D1-D2 diff: 5.152624551235657 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=95), RecessiveEncoder(), RecessiveEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17420405453792742 Holdout data R^2 trained on entire dataset(80%): 0.16105279269250472 Dataset D1 R^2 on trained D1: 0.1541735927145671 .................................................. Pipeline #21: Score on D2: 0.10954652987523339 | D1-D2 diff: 7.5545550311315575 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=5, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11933339437597246 Holdout data R^2 trained on entire dataset(80%): 0.10274328694767043 Dataset D1 R^2 on trained D1: 0.10985354830299876 .................................................. Pipeline #22: Score on D2: 0.03398295607326385 | D1-D2 diff: 8.349922600720577 Pipeline steps: DominantEncoder(), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.036114343790395576 Holdout data R^2 trained on entire dataset(80%): 0.009795747710299452 Dataset D1 R^2 on trained D1: 0.03377723906455954 .................................................. Pipeline #23: Score on D2: 0.02391372483777665 | D1-D2 diff: 21.149348332694633 Pipeline steps: DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.024048410265911535 Holdout data R^2 trained on entire dataset(80%): -0.004967269379323902 Dataset D1 R^2 on trained D1: 0.023918723019468513 .................................................. ************************************************************************************** Random Seed 24 - 8 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.18359 Best score on D1-D2 diff: 5.43747 Gen 2 - Best score on D2: 0.19165 Best score on D1-D2 diff: 5.43747 Gen 3 - Best score on D2: 0.19165 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.19334 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.19334 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.19334 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.19578 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.19578 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.19578 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.19618 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.19618 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.19636 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.19799 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.19799 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.19949 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.20076 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.20076 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.20076 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.20076 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.006784770835390086 Entire dataset(80%) R^2 trained on data (80%): 0.005181601675197989 Holdout R^2 (20%) trained on data (80%): 0.002639624522640105 Dataset D1 R^2 on trained D1: 0.008570000565647606 Combined Dataset (100%) R^2 trained on combined data (100%): 0.005676142750101865 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.2007622627503014 | D1-D2 diff: 1.4293480410388635 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=3, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4358734496497747 Holdout data R^2 trained on entire dataset(80%): 0.19951506858742118 Dataset D1 R^2 on trained D1: 0.44034087052273474 .................................................. Pipeline #2: Score on D2: 0.1957779224790378 | D1-D2 diff: 1.5748469523213204 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.36594827790555573 Holdout data R^2 trained on entire dataset(80%): 0.20261058292995449 Dataset D1 R^2 on trained D1: 0.35835023473336414 .................................................. Pipeline #3: Score on D2: 0.1931397288432547 | D1-D2 diff: 1.6040365409212676 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=3, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.349279764380131 Holdout data R^2 trained on entire dataset(80%): 0.20221640683839082 Dataset D1 R^2 on trained D1: 0.3441974642994585 .................................................. Pipeline #4: Score on D2: 0.1927567238381972 | D1-D2 diff: 1.6576256672419976 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=3, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3282575867462383 Holdout data R^2 trained on entire dataset(80%): 0.2013984103434988 Dataset D1 R^2 on trained D1: 0.32520739047214287 .................................................. Pipeline #5: Score on D2: 0.18769475162255578 | D1-D2 diff: 1.7170853014960483 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=15, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30821713319439326 Holdout data R^2 trained on entire dataset(80%): 0.1955387316299929 Dataset D1 R^2 on trained D1: 0.3027304193096443 .................................................. Pipeline #6: Score on D2: 0.1852568015805789 | D1-D2 diff: 1.7324058531863291 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=16, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30103000687929304 Holdout data R^2 trained on entire dataset(80%): 0.19294040728865336 Dataset D1 R^2 on trained D1: 0.2962768546125628 .................................................. Pipeline #7: Score on D2: 0.18501103083187576 | D1-D2 diff: 1.746843325860625 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=16, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2996229530624793 Holdout data R^2 trained on entire dataset(80%): 0.19325036058292355 Dataset D1 R^2 on trained D1: 0.2924060600094338 .................................................. Pipeline #8: Score on D2: 0.18425993533373053 | D1-D2 diff: 1.7472024062526006 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.295961048698393 Holdout data R^2 trained on entire dataset(80%): 0.19208425267707974 Dataset D1 R^2 on trained D1: 0.2915667055609167 .................................................. Pipeline #9: Score on D2: 0.18358753584399745 | D1-D2 diff: 1.7718026701440668 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29393326181542134 Holdout data R^2 trained on entire dataset(80%): 0.19049281213614555 Dataset D1 R^2 on trained D1: 0.2850577536616503 .................................................. Pipeline #10: Score on D2: 0.18174876524570693 | D1-D2 diff: 1.8486415330560515 Pipeline steps: VarianceThreshold(threshold=0.1), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2758404358958193 Holdout data R^2 trained on entire dataset(80%): 0.19067815442904956 Dataset D1 R^2 on trained D1: 0.2673715159753193 .................................................. Pipeline #11: Score on D2: 0.17943450343125023 | D1-D2 diff: 1.8670507457246306 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=14, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2712133153355375 Holdout data R^2 trained on entire dataset(80%): 0.1904639977263598 Dataset D1 R^2 on trained D1: 0.26172989419363557 .................................................. Pipeline #12: Score on D2: 0.17881694147497307 | D1-D2 diff: 1.9457534643355874 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2575297752921625 Holdout data R^2 trained on entire dataset(80%): 0.18887052694259043 Dataset D1 R^2 on trained D1: 0.2485837329490016 .................................................. Pipeline #13: Score on D2: 0.17233591680547677 | D1-D2 diff: 2.0224151965491513 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24205074719158737 Holdout data R^2 trained on entire dataset(80%): 0.18007953906232854 Dataset D1 R^2 on trained D1: 0.2321107978176704 .................................................. Pipeline #14: Score on D2: 0.1586718317964474 | D1-D2 diff: 2.208774935753636 Pipeline steps: SelectPercentile(percentile=60), FeatureEncodingFrequencySelector(threshold=0.25), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=16, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18846519136450246 Holdout data R^2 trained on entire dataset(80%): 0.14831306191085658 Dataset D1 R^2 on trained D1: 0.2006858421499076 .................................................. Pipeline #15: Score on D2: 0.14755832871914887 | D1-D2 diff: 2.3189634637441694 Pipeline steps: SelectPercentile(percentile=85), HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18963583755211777 Holdout data R^2 trained on entire dataset(80%): 0.15076636896765594 Dataset D1 R^2 on trained D1: 0.18213827985205888 .................................................. Pipeline #16: Score on D2: 0.14705224942523243 | D1-D2 diff: 2.356575713896929 Pipeline steps: SelectPercentile(percentile=80), HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19656610080772474 Holdout data R^2 trained on entire dataset(80%): 0.15628412556456428 Dataset D1 R^2 on trained D1: 0.17947683315829044 .................................................. Pipeline #17: Score on D2: 0.13227779726968403 | D1-D2 diff: 2.4453312270534755 Pipeline steps: VarianceThreshold(threshold=0.2), SelectPercentile(percentile=85), HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1698206753692737 Holdout data R^2 trained on entire dataset(80%): 0.1352577287000123 Dataset D1 R^2 on trained D1: 0.16024501249171152 .................................................. Pipeline #18: Score on D2: 0.12742626044601435 | D1-D2 diff: 2.4557289474363984 Pipeline steps: SelectPercentile(percentile=85), FeatureEncodingFrequencySelector(threshold=0.05), HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=20, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15362400056519498 Holdout data R^2 trained on entire dataset(80%): 0.11057913976591094 Dataset D1 R^2 on trained D1: 0.15492281522035545 .................................................. Pipeline #19: Score on D2: 0.11556919424142742 | D1-D2 diff: 4.449521833429415 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=85), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=12, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11784966971995381 Holdout data R^2 trained on entire dataset(80%): 0.11551507090131519 Dataset D1 R^2 on trained D1: 0.11812040672523871 .................................................. Pipeline #20: Score on D2: 0.10347188593305545 | D1-D2 diff: 5.7575058625098015 Pipeline steps: VarianceThreshold(threshold=0.15), VarianceThreshold(threshold=0.05), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10643970808101177 Holdout data R^2 trained on entire dataset(80%): 0.09704179853483397 Dataset D1 R^2 on trained D1: 0.10438193004605567 .................................................. Pipeline #21: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=5), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=3, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 24 - 9 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.17836 Best score on D1-D2 diff: 4.56097 Gen 2 - Best score on D2: 0.17968 Best score on D1-D2 diff: 5.64049 Gen 3 - Best score on D2: 0.17968 Best score on D1-D2 diff: 5.66950 Gen 4 - Best score on D2: 0.18046 Best score on D1-D2 diff: 5.68030 Gen 5 - Best score on D2: 0.18620 Best score on D1-D2 diff: 7.47566 Gen 6 - Best score on D2: 0.18620 Best score on D1-D2 diff: 7.47566 Gen 7 - Best score on D2: 0.18698 Best score on D1-D2 diff: 12.36062 Gen 8 - Best score on D2: 0.19124 Best score on D1-D2 diff: 12.36062 Gen 9 - Best score on D2: 0.19124 Best score on D1-D2 diff: 12.36062 Gen 10 - Best score on D2: 0.19124 Best score on D1-D2 diff: 12.36062 Gen 11 - Best score on D2: 0.19124 Best score on D1-D2 diff: 12.36062 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.009560919033643644 Entire dataset(80%) R^2 trained on data (80%): 0.0063484056825678925 Holdout R^2 (20%) trained on data (80%): -0.0065175446844278895 Dataset D1 R^2 on trained D1: 0.008694627170112823 Combined Dataset (100%) R^2 trained on combined data (100%): 0.005431457702462561 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.19123897664752076 | D1-D2 diff: 1.3511473860165584 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=2, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.48575733596713544 Holdout data R^2 trained on entire dataset(80%): 0.22086531670539855 Dataset D1 R^2 on trained D1: 0.49128584515268703 .................................................. Pipeline #2: Score on D2: 0.19042941467719243 | D1-D2 diff: 1.363044405537824 Pipeline steps: VarianceThreshold(threshold=0.05), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=2, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.47292202131156813 Holdout data R^2 trained on entire dataset(80%): 0.2189712032337786 Dataset D1 R^2 on trained D1: 0.4801370768837965 .................................................. Pipeline #3: Score on D2: 0.18726813774821272 | D1-D2 diff: 1.4796640084393378 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=2, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4005898260591757 Holdout data R^2 trained on entire dataset(80%): 0.22528521601147655 Dataset D1 R^2 on trained D1: 0.395884111634329 .................................................. Pipeline #4: Score on D2: 0.18620445595734436 | D1-D2 diff: 1.493104906532353 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=7, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3862767236666934 Holdout data R^2 trained on entire dataset(80%): 0.22461331420344177 Dataset D1 R^2 on trained D1: 0.38740942848062276 .................................................. Pipeline #5: Score on D2: 0.18374767891553756 | D1-D2 diff: 1.5944892659561838 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=11, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34262596678241863 Holdout data R^2 trained on entire dataset(80%): 0.22028698204175012 Dataset D1 R^2 on trained D1: 0.3384559740429035 .................................................. Pipeline #6: Score on D2: 0.1829994566670149 | D1-D2 diff: 1.600408101045285 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=3, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3474607259653577 Holdout data R^2 trained on entire dataset(80%): 0.22345811348169575 Dataset D1 R^2 on trained D1: 0.33543176831667054 .................................................. Pipeline #7: Score on D2: 0.18155460057043138 | D1-D2 diff: 1.636528641063504 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=11, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3264286181742362 Holdout data R^2 trained on entire dataset(80%): 0.22119425967045947 Dataset D1 R^2 on trained D1: 0.32096833434625205 .................................................. Pipeline #8: Score on D2: 0.17875164884226113 | D1-D2 diff: 1.660296108367781 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=6, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.317795786111173 Holdout data R^2 trained on entire dataset(80%): 0.2141705029470794 Dataset D1 R^2 on trained D1: 0.3103522278977634 .................................................. Pipeline #9: Score on D2: 0.17844246064643154 | D1-D2 diff: 1.733987374426751 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=12, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2970071761718298 Holdout data R^2 trained on entire dataset(80%): 0.22106396451526422 Dataset D1 R^2 on trained D1: 0.28905803436691535 .................................................. Pipeline #10: Score on D2: 0.17835939757651298 | D1-D2 diff: 1.7463274756926874 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29493864059902863 Holdout data R^2 trained on entire dataset(80%): 0.21728922670690765 Dataset D1 R^2 on trained D1: 0.28588137727476526 .................................................. Pipeline #11: Score on D2: 0.17496710245870284 | D1-D2 diff: 1.7974787585111753 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=11, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28133315841970286 Holdout data R^2 trained on entire dataset(80%): 0.21841016123605728 Dataset D1 R^2 on trained D1: 0.27076256363811857 .................................................. Pipeline #12: Score on D2: 0.1745999642795908 | D1-D2 diff: 1.870781431203103 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=17, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2661601477029123 Holdout data R^2 trained on entire dataset(80%): 0.21410140157923874 Dataset D1 R^2 on trained D1: 0.2562408669066937 .................................................. Pipeline #13: Score on D2: 0.17101128484624462 | D1-D2 diff: 1.9029357611679882 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2576762873632952 Holdout data R^2 trained on entire dataset(80%): 0.2128786385817134 Dataset D1 R^2 on trained D1: 0.24727245926099783 .................................................. Pipeline #14: Score on D2: 0.16195153875877066 | D1-D2 diff: 2.0191527205905166 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23489800767885438 Holdout data R^2 trained on entire dataset(80%): 0.2101025443347163 Dataset D1 R^2 on trained D1: 0.2221136857067878 .................................................. Pipeline #15: Score on D2: 0.1455397442302383 | D1-D2 diff: 2.0913439651431407 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.05, min_samples_leaf=17, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21695958406825921 Holdout data R^2 trained on entire dataset(80%): 0.19068503801448122 Dataset D1 R^2 on trained D1: 0.19781523636661102 .................................................. Pipeline #16: Score on D2: 0.13069488435169552 | D1-D2 diff: 2.1846661676535724 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=15, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19352791101459454 Holdout data R^2 trained on entire dataset(80%): 0.1798642566317944 Dataset D1 R^2 on trained D1: 0.17459439382161168 .................................................. Pipeline #17: Score on D2: 0.12317319003522775 | D1-D2 diff: 2.23868901256195 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1855640754970871 Holdout data R^2 trained on entire dataset(80%): 0.17668533621762217 Dataset D1 R^2 on trained D1: 0.16298619233318845 .................................................. Pipeline #18: Score on D2: 0.12190304168480115 | D1-D2 diff: 2.6750411741706186 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), DecisionTreeRegressor(max_depth=3, min_samples_leaf=12, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1326972164751843 Holdout data R^2 trained on entire dataset(80%): 0.13456653501385984 Dataset D1 R^2 on trained D1: 0.14143195671255304 .................................................. Pipeline #19: Score on D2: 0.09635795786965717 | D1-D2 diff: 2.917674088978856 Pipeline steps: DecisionTreeRegressor(max_depth=3, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11043199353590127 Holdout data R^2 trained on entire dataset(80%): 0.11670440302161778 Dataset D1 R^2 on trained D1: 0.1101571253520992 .................................................. Pipeline #20: Score on D2: 0.06731641303846436 | D1-D2 diff: 4.0054177229446 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07033728470358325 Holdout data R^2 trained on entire dataset(80%): 0.08466346659021329 Dataset D1 R^2 on trained D1: 0.06343125455289589 .................................................. Pipeline #21: Score on D2: 0.058180756945053425 | D1-D2 diff: 4.5609681154702715 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=15), DecisionTreeRegressor(max_depth=1, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.059703633035496306 Holdout data R^2 trained on entire dataset(80%): 0.06984283073310482 Dataset D1 R^2 on trained D1: 0.06049160747353599 .................................................. Pipeline #22: Score on D2: 0.01651731751720198 | D1-D2 diff: 6.030065082825761 Pipeline steps: SelectPercentile(percentile=60), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), UnderDominanceEncoder(), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09455813641076771 Holdout data R^2 trained on entire dataset(80%): 0.12071839803933582 Dataset D1 R^2 on trained D1: 0.015760986342106542 .................................................. Pipeline #23: Score on D2: 0.014168890981530291 | D1-D2 diff: 7.527468984303493 Pipeline steps: SelectPercentile(percentile=35), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05970363303549653 Holdout data R^2 trained on entire dataset(80%): 0.06984283073310482 Dataset D1 R^2 on trained D1: 0.01448035228801725 .................................................. Pipeline #24: Score on D2: 0.0038488958488750447 | D1-D2 diff: 12.360624046451834 Pipeline steps: SelectPercentile(percentile=50), RecessiveEncoder(), VarianceThreshold(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.15), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03560538668585678 Holdout data R^2 trained on entire dataset(80%): 0.03297696769715064 Dataset D1 R^2 on trained D1: 0.003891734758721621 .................................................. ************************************************************************************** Random Seed 25 - 0 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00218 Best score on D1-D2 diff: 3.72463 Gen 2 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 3 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.006992146977758518 Entire dataset(80%) R^2 trained on data (80%): 0.003948346076884102 Holdout R^2 (20%) trained on data (80%): -0.00398263365221152 Dataset D1 R^2 on trained D1: 0.008656781061245455 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0032637296202961963 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: -3.2015594250811574e-05 | D1-D2 diff: 13.480151627526027 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=9, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021467281316e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516763899763 Dataset D1 R^2 on trained D1: -1.73106056000627e-06 .................................................. ************************************************************************************** Random Seed 25 - 1 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.08909 Best score on D1-D2 diff: 4.57112 Gen 2 - Best score on D2: 0.08909 Best score on D1-D2 diff: 4.57112 Gen 3 - Best score on D2: 0.08909 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.08909 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.08909 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.09074 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.09199 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 10 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 11 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 12 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 13 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 14 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 15 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 16 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 17 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 18 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 19 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 20 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 Gen 21 - Best score on D2: 0.09571 Best score on D1-D2 diff: 13.95341 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.007202488064233403 Entire dataset(80%) R^2 trained on data (80%): 0.0035230986559424693 Holdout R^2 (20%) trained on data (80%): -0.0033110281509347583 Dataset D1 R^2 on trained D1: 0.007898259779421823 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0031172331237309114 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.09571355976467144 | D1-D2 diff: 3.071335578031102 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09051408479496947 Holdout data R^2 trained on entire dataset(80%): 0.11957740250877347 Dataset D1 R^2 on trained D1: 0.08447551086248939 .................................................. Pipeline #2: Score on D2: 0.09571355976467133 | D1-D2 diff: 3.0713355780311096 Pipeline steps: SelectPercentile(percentile=35), OverDominanceEncoder(), VarianceThreshold(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=13, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0014526224734418003 Holdout data R^2 trained on entire dataset(80%): -0.001601609368541368 Dataset D1 R^2 on trained D1: 0.08447551086248939 .................................................. Pipeline #3: Score on D2: 0.09200638319712973 | D1-D2 diff: 4.938095745782655 Pipeline steps: SelectPercentile(percentile=45), VarianceThreshold(threshold=0.25), VarianceThreshold(), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0023114485221359393 Holdout data R^2 trained on entire dataset(80%): -0.004183820815817363 Dataset D1 R^2 on trained D1: 0.09032463111214717 .................................................. Pipeline #4: Score on D2: 0.09073955999128813 | D1-D2 diff: 11.042579582642043 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=8, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00394850040262007 Holdout data R^2 trained on entire dataset(80%): -0.0008767194243173293 Dataset D1 R^2 on trained D1: 0.09080681394932733 .................................................. Pipeline #5: Score on D2: -2.4080840544948856e-05 | D1-D2 diff: 13.953408795943623 Pipeline steps: SelectPercentile(percentile=10), FeatureEncodingFrequencySelector(threshold=0.05), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0009136197755612585 Holdout data R^2 trained on entire dataset(80%): -0.0002353779529919997 Dataset D1 R^2 on trained D1: 2.2993985770991543e-06 .................................................. ************************************************************************************** Random Seed 25 - 2 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.07274 Best score on D1-D2 diff: 3.72463 Gen 2 - Best score on D2: 0.08026 Best score on D1-D2 diff: 13.48015 Gen 3 - Best score on D2: 0.08425 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.08425 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.09147 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.09147 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.09147 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.004967991161772822 Entire dataset(80%) R^2 trained on data (80%): 0.0038466440187252537 Holdout R^2 (20%) trained on data (80%): -0.004709535776275642 Dataset D1 R^2 on trained D1: 0.007354930635341539 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0029971666854403667 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.09147077975158158 | D1-D2 diff: 1.797733578328369 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19261541364815793 Holdout data R^2 trained on entire dataset(80%): 0.13739170102989384 Dataset D1 R^2 on trained D1: 0.18721193835230154 .................................................. Pipeline #2: Score on D2: 0.08651469081831986 | D1-D2 diff: 1.7982361788285637 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19358375634356773 Holdout data R^2 trained on entire dataset(80%): 0.13547837508366856 Dataset D1 R^2 on trained D1: 0.18214885705693462 .................................................. Pipeline #3: Score on D2: 0.08624033569633227 | D1-D2 diff: 1.8583422141037411 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17688414527764595 Holdout data R^2 trained on entire dataset(80%): 0.12788134057908385 Dataset D1 R^2 on trained D1: 0.170089208800713 .................................................. Pipeline #4: Score on D2: 0.05914585397274952 | D1-D2 diff: 2.082164560884644 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=5, min_samples_leaf=20, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11496833578334242 Holdout data R^2 trained on entire dataset(80%): 0.09680440205008267 Dataset D1 R^2 on trained D1: 0.11234930437685686 .................................................. Pipeline #5: Score on D2: 0.056699105323649523 | D1-D2 diff: 3.09520828824231 Pipeline steps: VarianceThreshold(), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=7, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10131789639947064 Holdout data R^2 trained on entire dataset(80%): 0.11121578767915108 Dataset D1 R^2 on trained D1: 0.06759443764683226 .................................................. Pipeline #6: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: DominantEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=9, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 25 - 3 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.10639 Best score on D1-D2 diff: 4.07963 Gen 2 - Best score on D2: 0.11181 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.11288 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.11288 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.11288 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.11288 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.11288 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.11288 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.11288 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.11288 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.11288 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.11288 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.11446 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.11446 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.11446 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.11446 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.004591361329808308 Entire dataset(80%) R^2 trained on data (80%): 0.005300261647207982 Holdout R^2 (20%) trained on data (80%): -0.0007759414932348996 Dataset D1 R^2 on trained D1: 0.008829243408843168 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004848892336837052 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.11446228388873436 | D1-D2 diff: 1.7785214772236437 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=17, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2232829241995804 Holdout data R^2 trained on entire dataset(80%): 0.1707850402028388 Dataset D1 R^2 on trained D1: 0.21440785266096163 .................................................. Pipeline #2: Score on D2: 0.11258456140165873 | D1-D2 diff: 1.8041719252091442 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21747060314409916 Holdout data R^2 trained on entire dataset(80%): 0.1719860204743946 Dataset D1 R^2 on trained D1: 0.20696637496821269 .................................................. Pipeline #3: Score on D2: 0.10838723700203168 | D1-D2 diff: 1.8141617856768912 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21734303409370748 Holdout data R^2 trained on entire dataset(80%): 0.16690976460974538 Dataset D1 R^2 on trained D1: 0.20070726805354266 .................................................. Pipeline #4: Score on D2: 0.10562968275897333 | D1-D2 diff: 2.6747455726835083 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11425303909911388 Holdout data R^2 trained on entire dataset(80%): 0.1374903925576696 Dataset D1 R^2 on trained D1: 0.125167232227963 .................................................. Pipeline #5: Score on D2: 0.08301320768066245 | D1-D2 diff: 3.0299626102789996 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.2), DecisionTreeRegressor(max_depth=2, min_samples_leaf=16, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09056373424332365 Holdout data R^2 trained on entire dataset(80%): 0.11896988290925392 Dataset D1 R^2 on trained D1: 0.09487774816619687 .................................................. Pipeline #6: Score on D2: 0.02338068597173759 | D1-D2 diff: 3.1682769460957756 Pipeline steps: VarianceThreshold(threshold=0.35), SelectPercentile(percentile=50), DecisionTreeRegressor(max_depth=3, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07310265755681 Holdout data R^2 trained on entire dataset(80%): 0.08471973111602049 Dataset D1 R^2 on trained D1: 0.033305158893497167 .................................................. Pipeline #7: Score on D2: 0.02291456520081514 | D1-D2 diff: 5.5620269247417875 Pipeline steps: VarianceThreshold(threshold=0.2), DominantEncoder(), SelectPercentile(percentile=50), DecisionTreeRegressor(max_depth=3, min_samples_leaf=7, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.02483241464081254 Holdout data R^2 trained on entire dataset(80%): 0.03714219326151902 Dataset D1 R^2 on trained D1: 0.02395944817469642 .................................................. Pipeline #8: Score on D2: 0.001030224372324251 | D1-D2 diff: 6.120813616266627 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=35), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0019904022508646646 Holdout data R^2 trained on entire dataset(80%): 0.003364035762190576 Dataset D1 R^2 on trained D1: 0.0017426891185053917 .................................................. Pipeline #9: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: DominantEncoder(), DominantEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 25 - 4 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.12137 Best score on D1-D2 diff: 4.45439 Gen 2 - Best score on D2: 0.12409 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.12538 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.12545 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.12545 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.12805 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.12841 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.12841 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.008066717910654386 Entire dataset(80%) R^2 trained on data (80%): 0.006796087915162952 Holdout R^2 (20%) trained on data (80%): -0.004626101076089295 Dataset D1 R^2 on trained D1: 0.014474642160288376 Combined Dataset (100%) R^2 trained on combined data (100%): 0.005419582617260943 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1284063987983013 | D1-D2 diff: 1.5413061585392345 Pipeline steps: VarianceThreshold(threshold=0.35), UnderDominanceEncoder(), RecessiveEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.296220554061446 Holdout data R^2 trained on entire dataset(80%): 0.18672842997077055 Dataset D1 R^2 on trained D1: 0.30559849469340494 .................................................. Pipeline #2: Score on D2: 0.12805064411244327 | D1-D2 diff: 1.7828246401667087 Pipeline steps: VarianceThreshold(threshold=0.35), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2276061818765125 Holdout data R^2 trained on entire dataset(80%): 0.1775008945882982 Dataset D1 R^2 on trained D1: 0.22703475555801844 .................................................. Pipeline #3: Score on D2: 0.1258034385115614 | D1-D2 diff: 1.7911981374415324 Pipeline steps: VarianceThreshold(threshold=0.35), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=16, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22234370960960514 Holdout data R^2 trained on entire dataset(80%): 0.17494674783186825 Dataset D1 R^2 on trained D1: 0.22294956398497212 .................................................. Pipeline #4: Score on D2: 0.1255091910778271 | D1-D2 diff: 1.8171793996185002 Pipeline steps: VarianceThreshold(threshold=0.35), VarianceThreshold(threshold=0.1), UnderDominanceEncoder(), RecessiveEncoder(), VarianceThreshold(threshold=0.05), RandomForestRegressor(max_features=0.4, min_samples_leaf=17, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21829988605606188 Holdout data R^2 trained on entire dataset(80%): 0.17295097483749067 Dataset D1 R^2 on trained D1: 0.21721752017184914 .................................................. Pipeline #5: Score on D2: 0.11741271778380369 | D1-D2 diff: 1.8217158174269024 Pipeline steps: VarianceThreshold(threshold=0.35), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=16, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2113166534180183 Holdout data R^2 trained on entire dataset(80%): 0.16925536762405535 Dataset D1 R^2 on trained D1: 0.20821096882672396 .................................................. Pipeline #6: Score on D2: 0.11527483072418088 | D1-D2 diff: 1.8374932625030915 Pipeline steps: VarianceThreshold(threshold=0.25), UnderDominanceEncoder(), RecessiveEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=16, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20893709780267422 Holdout data R^2 trained on entire dataset(80%): 0.16459668388933024 Dataset D1 R^2 on trained D1: 0.20299449859056107 .................................................. Pipeline #7: Score on D2: 0.11448060152883421 | D1-D2 diff: 1.8534291631861495 Pipeline steps: VarianceThreshold(threshold=0.35), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20492592170688162 Holdout data R^2 trained on entire dataset(80%): 0.16696078267202719 Dataset D1 R^2 on trained D1: 0.19922207887622312 .................................................. Pipeline #8: Score on D2: 0.10468060955771341 | D1-D2 diff: 1.8841792033737432 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18868476556345815 Holdout data R^2 trained on entire dataset(80%): 0.15434519711689298 Dataset D1 R^2 on trained D1: 0.18402407703894186 .................................................. Pipeline #9: Score on D2: 0.10338659751486512 | D1-D2 diff: 1.9728261302291534 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1801760429184326 Holdout data R^2 trained on entire dataset(80%): 0.153200558272639 Dataset D1 R^2 on trained D1: 0.16940192070464255 .................................................. Pipeline #10: Score on D2: 0.09626885883297331 | D1-D2 diff: 2.0090124558446267 Pipeline steps: VarianceThreshold(threshold=0.35), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16866596593572158 Holdout data R^2 trained on entire dataset(80%): 0.14363648178905564 Dataset D1 R^2 on trained D1: 0.1576548796743904 .................................................. Pipeline #11: Score on D2: 0.09494012379692596 | D1-D2 diff: 2.023268927331291 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16510661706145024 Holdout data R^2 trained on entire dataset(80%): 0.13866071273677494 Dataset D1 R^2 on trained D1: 0.15461417913116282 .................................................. Pipeline #12: Score on D2: 0.08861657978318682 | D1-D2 diff: 2.082394391989978 Pipeline steps: SelectPercentile(percentile=85), VarianceThreshold(threshold=0.25), OverDominanceEncoder(), DominantEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11987657758395431 Holdout data R^2 trained on entire dataset(80%): 0.1136518823949072 Dataset D1 R^2 on trained D1: 0.14179654609867087 .................................................. Pipeline #13: Score on D2: 0.08325915175327736 | D1-D2 diff: 2.1093511112104752 Pipeline steps: VarianceThreshold(threshold=0.35), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14592543397629631 Holdout data R^2 trained on entire dataset(80%): 0.12276507927373759 Dataset D1 R^2 on trained D1: 0.13377230684111197 .................................................. Pipeline #14: Score on D2: 0.07816315912360361 | D1-D2 diff: 2.2788520930987546 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=10, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09831317000572126 Holdout data R^2 trained on entire dataset(80%): 0.09277122418022832 Dataset D1 R^2 on trained D1: 0.11524279371334356 .................................................. Pipeline #15: Score on D2: 0.04059941176395754 | D1-D2 diff: 2.2935920902945592 Pipeline steps: VarianceThreshold(threshold=0.35), UnderDominanceEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08659435815451755 Holdout data R^2 trained on entire dataset(80%): 0.05744436477941173 Dataset D1 R^2 on trained D1: 0.07673501183819575 .................................................. Pipeline #16: Score on D2: 0.03878819596489924 | D1-D2 diff: 2.334919830724624 Pipeline steps: VarianceThreshold(threshold=0.35), DominantEncoder(), UnderDominanceEncoder(), DominantEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08088397748061815 Holdout data R^2 trained on entire dataset(80%): 0.050225877593807144 Dataset D1 R^2 on trained D1: 0.07243254309549918 .................................................. Pipeline #17: Score on D2: 0.036356467945200754 | D1-D2 diff: 2.337108343233543 Pipeline steps: VarianceThreshold(threshold=0.35), DominantEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07827586231383876 Holdout data R^2 trained on entire dataset(80%): 0.049445530682629024 Dataset D1 R^2 on trained D1: 0.06987497116272001 .................................................. Pipeline #18: Score on D2: 0.0361324258017901 | D1-D2 diff: 2.360942111153231 Pipeline steps: VarianceThreshold(threshold=0.35), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), UnderDominanceEncoder(), DominantEncoder(), VarianceThreshold(threshold=0.1), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07998160757057826 Holdout data R^2 trained on entire dataset(80%): 0.05204272859121495 Dataset D1 R^2 on trained D1: 0.06831780614701122 .................................................. Pipeline #19: Score on D2: 0.035173328526001746 | D1-D2 diff: 2.364205512261389 Pipeline steps: VarianceThreshold(threshold=0.35), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07689205134856913 Holdout data R^2 trained on entire dataset(80%): 0.047093759521904976 Dataset D1 R^2 on trained D1: 0.0671813697485909 .................................................. Pipeline #20: Score on D2: 0.034249552599447064 | D1-D2 diff: 2.3829748380018954 Pipeline steps: VarianceThreshold(threshold=0.35), DominantEncoder(), UnderDominanceEncoder(), OverDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0725099772063853 Holdout data R^2 trained on entire dataset(80%): 0.045428221960225224 Dataset D1 R^2 on trained D1: 0.06526100977684801 .................................................. Pipeline #21: Score on D2: 0.03407062089375634 | D1-D2 diff: 2.4015287964353735 Pipeline steps: VarianceThreshold(threshold=0.35), DominantEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=95), VarianceThreshold(threshold=0.05), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07465739540980154 Holdout data R^2 trained on entire dataset(80%): 0.045351285501967764 Dataset D1 R^2 on trained D1: 0.06413476231593329 .................................................. Pipeline #22: Score on D2: 0.029619690480298555 | D1-D2 diff: 2.541405685865332 Pipeline steps: VarianceThreshold(threshold=0.25), DominantEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04592545183066454 Holdout data R^2 trained on entire dataset(80%): 0.022728893832990815 Dataset D1 R^2 on trained D1: 0.053591676242301434 .................................................. Pipeline #23: Score on D2: 0.02634008038981006 | D1-D2 diff: 2.7672311679680766 Pipeline steps: VarianceThreshold(threshold=0.05), SelectPercentile(percentile=70), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=9, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09831317000572126 Holdout data R^2 trained on entire dataset(80%): 0.09277122418022832 Dataset D1 R^2 on trained D1: 0.04339376542094642 .................................................. Pipeline #24: Score on D2: 0.02404786981152296 | D1-D2 diff: 3.6254409250326334 Pipeline steps: DominantEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.027197618991343764 Holdout data R^2 trained on entire dataset(80%): 0.025864205919307603 Dataset D1 R^2 on trained D1: 0.02983624493589465 .................................................. Pipeline #25: Score on D2: 0.0021178465689465353 | D1-D2 diff: 4.454386328980225 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0035134433662064035 Holdout data R^2 trained on entire dataset(80%): 0.004291630571857441 Dataset D1 R^2 on trained D1: 0.004657932898708084 .................................................. Pipeline #26: Score on D2: 0.0006295844363282121 | D1-D2 diff: 6.685774373682467 Pipeline steps: SelectPercentile(percentile=20), OverDominanceEncoder(), SelectPercentile(percentile=15), VarianceThreshold(threshold=0.05), HeterosisEncoder(), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0027771698596190664 Holdout data R^2 trained on entire dataset(80%): -0.003265637494964002 Dataset D1 R^2 on trained D1: 0.00012909705185681908 .................................................. Pipeline #27: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 25 - 5 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.14130 Best score on D1-D2 diff: 3.89121 Gen 2 - Best score on D2: 0.14130 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14130 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.14130 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.14130 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.14221 Best score on D1-D2 diff: 11.96036 Gen 7 - Best score on D2: 0.14221 Best score on D1-D2 diff: 11.96036 Gen 8 - Best score on D2: 0.14278 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.14278 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.14282 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.14282 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.14282 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.14282 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.14282 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.14282 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.14282 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.14457 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.14479 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.14479 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.14479 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.14848 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.14848 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.14848 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.14848 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.14848 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.009830037629097177 Entire dataset(80%) R^2 trained on data (80%): 0.0046915572047000476 Holdout R^2 (20%) trained on data (80%): -0.004731185920161973 Dataset D1 R^2 on trained D1: 0.011784079252599722 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004000247654956879 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.14848378522515027 | D1-D2 diff: 1.3423597873432347 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=2, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4489004648115954 Holdout data R^2 trained on entire dataset(80%): 0.20107066313840527 Dataset D1 R^2 on trained D1: 0.4564650267646694 .................................................. Pipeline #2: Score on D2: 0.14791022652352592 | D1-D2 diff: 1.4370097347036919 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3813340461079038 Holdout data R^2 trained on entire dataset(80%): 0.20052904815010242 Dataset D1 R^2 on trained D1: 0.38242011454261926 .................................................. Pipeline #3: Score on D2: 0.1461533084504767 | D1-D2 diff: 1.4640847820427032 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=6, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3612495434143319 Holdout data R^2 trained on entire dataset(80%): 0.2005804890228453 Dataset D1 R^2 on trained D1: 0.36379149268528455 .................................................. Pipeline #4: Score on D2: 0.1447890397760001 | D1-D2 diff: 1.627585744736395 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=11, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28590342535093816 Holdout data R^2 trained on entire dataset(80%): 0.19523218218739125 Dataset D1 R^2 on trained D1: 0.28729219827659036 .................................................. Pipeline #5: Score on D2: 0.14456780264709934 | D1-D2 diff: 1.767997140424414 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24630816620597784 Holdout data R^2 trained on entire dataset(80%): 0.19199306223533874 Dataset D1 R^2 on trained D1: 0.24691448447768516 .................................................. Pipeline #6: Score on D2: 0.14217294643385148 | D1-D2 diff: 1.7739188766155607 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24205575065622753 Holdout data R^2 trained on entire dataset(80%): 0.19261927545769697 Dataset D1 R^2 on trained D1: 0.24315983214101067 .................................................. Pipeline #7: Score on D2: 0.14132018445551564 | D1-D2 diff: 1.79051584485228 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2374764535231787 Holdout data R^2 trained on entire dataset(80%): 0.19094024506469387 Dataset D1 R^2 on trained D1: 0.2386144682988166 .................................................. Pipeline #8: Score on D2: 0.14007617998540633 | D1-D2 diff: 1.8062316493899515 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=20, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.233140289795438 Holdout data R^2 trained on entire dataset(80%): 0.18461945197885565 Dataset D1 R^2 on trained D1: 0.23402821870785573 .................................................. Pipeline #9: Score on D2: 0.13663095578565176 | D1-D2 diff: 1.8135174454961194 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22986385011846266 Holdout data R^2 trained on entire dataset(80%): 0.1875619810748056 Dataset D1 R^2 on trained D1: 0.22908226148429267 .................................................. Pipeline #10: Score on D2: 0.1362703785163517 | D1-D2 diff: 1.828724527604419 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22670884238786637 Holdout data R^2 trained on entire dataset(80%): 0.18456283874531176 Dataset D1 R^2 on trained D1: 0.2256846496123337 .................................................. Pipeline #11: Score on D2: 0.13573040680282833 | D1-D2 diff: 1.8336925679704483 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22772212477041998 Holdout data R^2 trained on entire dataset(80%): 0.1824203930246382 Dataset D1 R^2 on trained D1: 0.2241796051144177 .................................................. Pipeline #12: Score on D2: 0.12882408078079888 | D1-D2 diff: 1.8598266230253682 Pipeline steps: SelectPercentile(percentile=90), HeterosisEncoder(), OverDominanceEncoder(), VarianceThreshold(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22250540793038964 Holdout data R^2 trained on entire dataset(80%): 0.18833269836155253 Dataset D1 R^2 on trained D1: 0.21240558039909718 .................................................. Pipeline #13: Score on D2: 0.12820995268280622 | D1-D2 diff: 1.8743200157445945 Pipeline steps: SelectPercentile(percentile=90), HeterosisEncoder(), OverDominanceEncoder(), VarianceThreshold(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21910460721216274 Holdout data R^2 trained on entire dataset(80%): 0.18617760943801076 Dataset D1 R^2 on trained D1: 0.2092360698995792 .................................................. Pipeline #14: Score on D2: 0.121137464675521 | D1-D2 diff: 2.1151948623875376 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=6, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14434081148010525 Holdout data R^2 trained on entire dataset(80%): 0.13977541109522018 Dataset D1 R^2 on trained D1: 0.17109470829057838 .................................................. Pipeline #15: Score on D2: 0.10657573694663447 | D1-D2 diff: 3.49043752371144 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=10, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11027599426242918 Holdout data R^2 trained on entire dataset(80%): 0.12555399244394583 Dataset D1 R^2 on trained D1: 0.11331295377194761 .................................................. Pipeline #16: Score on D2: 0.06207094430707405 | D1-D2 diff: 3.7446597907737336 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06478972864326116 Holdout data R^2 trained on entire dataset(80%): 0.08471864577325139 Dataset D1 R^2 on trained D1: 0.06715664187683235 .................................................. Pipeline #17: Score on D2: 0.026316103798644352 | D1-D2 diff: 3.796992790754135 Pipeline steps: VarianceThreshold(), DominantEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.029071123630685203 Holdout data R^2 trained on entire dataset(80%): 0.027740135286987155 Dataset D1 R^2 on trained D1: 0.03112716531223836 .................................................. Pipeline #18: Score on D2: 0.01883565213445282 | D1-D2 diff: 3.8912055402405326 Pipeline steps: DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.021138376125381275 Holdout data R^2 trained on entire dataset(80%): 0.02229871647919135 Dataset D1 R^2 on trained D1: 0.02319742768330868 .................................................. Pipeline #19: Score on D2: 0.009901760648260871 | D1-D2 diff: 4.412101827242649 Pipeline steps: VarianceThreshold(threshold=0.15), DominantEncoder(), RecessiveEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01130305992592695 Holdout data R^2 trained on entire dataset(80%): 0.01701601579759171 Dataset D1 R^2 on trained D1: 0.01254063001708372 .................................................. Pipeline #20: Score on D2: 0.0003323477737967595 | D1-D2 diff: 8.137739860276785 Pipeline steps: VarianceThreshold(threshold=0.2), SelectPercentile(percentile=65), RecessiveEncoder(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0004723174671678043 Holdout data R^2 trained on entire dataset(80%): -0.0008971269046422758 Dataset D1 R^2 on trained D1: 0.0005603739946066222 .................................................. Pipeline #21: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=15, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 25 - 6 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.15062 Best score on D1-D2 diff: 3.89121 Gen 2 - Best score on D2: 0.16002 Best score on D1-D2 diff: 6.15897 Gen 3 - Best score on D2: 0.16002 Best score on D1-D2 diff: 6.15897 Gen 4 - Best score on D2: 0.16002 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.16002 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.16002 Best score on D1-D2 diff: 11.96036 Gen 7 - Best score on D2: 0.16002 Best score on D1-D2 diff: 11.96036 Gen 8 - Best score on D2: 0.16002 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.16002 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.16002 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.16002 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.16002 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.16002 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.16002 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.16105 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.16105 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.16259 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.16259 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.16259 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.16259 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.16259 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.16259 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.16259 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.16259 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.16259 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0068371157769921975 Entire dataset(80%) R^2 trained on data (80%): 0.004595631028720737 Holdout R^2 (20%) trained on data (80%): -0.006811449278390613 Dataset D1 R^2 on trained D1: 0.00997988061346411 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003680455931548199 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.16259007497931544 | D1-D2 diff: 1.492406178079701 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.35904148544838776 Holdout data R^2 trained on entire dataset(80%): 0.223507758192933 Dataset D1 R^2 on trained D1: 0.3641721201907807 .................................................. Pipeline #2: Score on D2: 0.1567823179589658 | D1-D2 diff: 1.5052331154757073 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=7, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34561561470592805 Holdout data R^2 trained on entire dataset(80%): 0.2103078913317753 Dataset D1 R^2 on trained D1: 0.3515805193661151 .................................................. Pipeline #3: Score on D2: 0.15673022269791204 | D1-D2 diff: 1.5176797885833648 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=7, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33619973221015687 Holdout data R^2 trained on entire dataset(80%): 0.21271949547760116 Dataset D1 R^2 on trained D1: 0.3452163528231814 .................................................. Pipeline #4: Score on D2: 0.1547863921305277 | D1-D2 diff: 1.7251113231810984 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=95), UnderDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2542614926595329 Holdout data R^2 trained on entire dataset(80%): 0.19356942566906132 Dataset D1 R^2 on trained D1: 0.267696155412233 .................................................. Pipeline #5: Score on D2: 0.15202406525219458 | D1-D2 diff: 1.760638425104736 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2537495260246013 Holdout data R^2 trained on entire dataset(80%): 0.20137086895148248 Dataset D1 R^2 on trained D1: 0.25609256562306437 .................................................. Pipeline #6: Score on D2: 0.15172671044064912 | D1-D2 diff: 1.7786157771309468 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=19, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2503717893964166 Holdout data R^2 trained on entire dataset(80%): 0.2030200390422665 Dataset D1 R^2 on trained D1: 0.25165108495950794 .................................................. Pipeline #7: Score on D2: 0.14853701606709835 | D1-D2 diff: 1.7822986654222124 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24683151470485543 Holdout data R^2 trained on entire dataset(80%): 0.19675585074395585 Dataset D1 R^2 on trained D1: 0.2476380241806193 .................................................. Pipeline #8: Score on D2: 0.14802599672140704 | D1-D2 diff: 1.8236313709169527 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2372550929901407 Holdout data R^2 trained on entire dataset(80%): 0.1979765114401183 Dataset D1 R^2 on trained D1: 0.23844334829675584 .................................................. Pipeline #9: Score on D2: 0.14479453325480962 | D1-D2 diff: 1.8525672004386646 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23138991882681603 Holdout data R^2 trained on entire dataset(80%): 0.19387987398686357 Dataset D1 R^2 on trained D1: 0.22969383481785022 .................................................. Pipeline #10: Score on D2: 0.1446605652981675 | D1-D2 diff: 1.862612991054353 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23034028651834637 Holdout data R^2 trained on entire dataset(80%): 0.19667351372540154 Dataset D1 R^2 on trained D1: 0.2277430524927826 .................................................. Pipeline #11: Score on D2: 0.1439094734053118 | D1-D2 diff: 1.8845133590882595 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22506194762193787 Holdout data R^2 trained on entire dataset(80%): 0.19432594556098537 Dataset D1 R^2 on trained D1: 0.22319668016128158 .................................................. Pipeline #12: Score on D2: 0.1412873730591715 | D1-D2 diff: 1.9207864543281645 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=95), RecessiveEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20814106006508504 Holdout data R^2 trained on entire dataset(80%): 0.18240432029148967 Dataset D1 R^2 on trained D1: 0.2147529109314813 .................................................. Pipeline #13: Score on D2: 0.13780316054992636 | D1-D2 diff: 1.9299667419298805 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21203975153614962 Holdout data R^2 trained on entire dataset(80%): 0.1888200727136935 Dataset D1 R^2 on trained D1: 0.20988082398856756 .................................................. Pipeline #14: Score on D2: 0.13707874895373995 | D1-D2 diff: 1.9342958014915264 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21478822831964028 Holdout data R^2 trained on entire dataset(80%): 0.1901282531823485 Dataset D1 R^2 on trained D1: 0.20851332035923553 .................................................. Pipeline #15: Score on D2: 0.1314429281442654 | D1-D2 diff: 2.4064203233608286 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13899268532786313 Holdout data R^2 trained on entire dataset(80%): 0.14481003048218688 Dataset D1 R^2 on trained D1: 0.16126336854455825 .................................................. Pipeline #16: Score on D2: 0.10657573694663447 | D1-D2 diff: 3.49043752371144 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=95), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11027599426242918 Holdout data R^2 trained on entire dataset(80%): 0.12555399244394583 Dataset D1 R^2 on trained D1: 0.11331295377194761 .................................................. Pipeline #17: Score on D2: 0.06207094430707405 | D1-D2 diff: 3.7446597907737336 Pipeline steps: SelectPercentile(percentile=85), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06478972864326116 Holdout data R^2 trained on entire dataset(80%): 0.08471864577325139 Dataset D1 R^2 on trained D1: 0.06715664187683235 .................................................. Pipeline #18: Score on D2: 0.026316103798644463 | D1-D2 diff: 3.796992790754157 Pipeline steps: DominantEncoder(), VarianceThreshold(threshold=0.05), DecisionTreeRegressor(max_depth=2, min_samples_leaf=10, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.029071123630685203 Holdout data R^2 trained on entire dataset(80%): 0.027740135286987155 Dataset D1 R^2 on trained D1: 0.03112716531223836 .................................................. Pipeline #19: Score on D2: 0.01883565213445282 | D1-D2 diff: 3.8912055402405326 Pipeline steps: DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.021138376125381275 Holdout data R^2 trained on entire dataset(80%): 0.02229871647919135 Dataset D1 R^2 on trained D1: 0.02319742768330868 .................................................. Pipeline #20: Score on D2: 0.009918083773003294 | D1-D2 diff: 4.420095711055364 Pipeline steps: DominantEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011302045229638957 Holdout data R^2 trained on entire dataset(80%): 0.01695638185912407 Dataset D1 R^2 on trained D1: 0.012537914949438567 .................................................. Pipeline #21: Score on D2: 0.0003323477737967595 | D1-D2 diff: 8.137739860276785 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00047231746716802636 Holdout data R^2 trained on entire dataset(80%): -0.0008971269046422758 Dataset D1 R^2 on trained D1: 0.0005603739946066222 .................................................. Pipeline #22: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=17, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 25 - 7 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.16399 Best score on D1-D2 diff: 3.79516 Gen 2 - Best score on D2: 0.16399 Best score on D1-D2 diff: 11.96036 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0045137532552945725 Entire dataset(80%) R^2 trained on data (80%): 0.004955481403276507 Holdout R^2 (20%) trained on data (80%): -0.006620778820909612 Dataset D1 R^2 on trained D1: 0.009521652265471081 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0036892982128127194 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.16399481557599305 | D1-D2 diff: 1.708090312569568 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2824154531775832 Holdout data R^2 trained on entire dataset(80%): 0.21513291483036345 Dataset D1 R^2 on trained D1: 0.28147285272844846 .................................................. Pipeline #2: Score on D2: 0.1347865281586781 | D1-D2 diff: 1.7588778945051413 Pipeline steps: RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2458776831830516 Holdout data R^2 trained on entire dataset(80%): 0.17668625226797496 Dataset D1 R^2 on trained D1: 0.23927231967963747 .................................................. Pipeline #3: Score on D2: 0.12928707586405597 | D1-D2 diff: 2.093591353999598 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=18, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18919330808835766 Holdout data R^2 trained on entire dataset(80%): 0.17215770161191124 Dataset D1 R^2 on trained D1: 0.1813384663429779 .................................................. Pipeline #4: Score on D2: 0.12806483510597788 | D1-D2 diff: 2.4068131805274295 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15236278376738155 Holdout data R^2 trained on entire dataset(80%): 0.15096092057655042 Dataset D1 R^2 on trained D1: 0.1578658102552436 .................................................. Pipeline #5: Score on D2: 0.10739059752111746 | D1-D2 diff: 3.640893747536217 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11046388777183413 Holdout data R^2 trained on entire dataset(80%): 0.12475473966971451 Dataset D1 R^2 on trained D1: 0.11308132750762523 .................................................. Pipeline #6: Score on D2: 0.05479124769642241 | D1-D2 diff: 3.9043535678661336 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=3, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06087298799980845 Holdout data R^2 trained on entire dataset(80%): 0.07972131182955433 Dataset D1 R^2 on trained D1: 0.05909456572211036 .................................................. Pipeline #7: Score on D2: -4.886780840163141e-05 | D1-D2 diff: 11.960360895673112 Pipeline steps: RecessiveEncoder(), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0 Holdout data R^2 trained on entire dataset(80%): -0.0012028184006112053 Dataset D1 R^2 on trained D1: 0.0 .................................................. ************************************************************************************** Random Seed 25 - 8 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.17584 Best score on D1-D2 diff: 9.49027 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.007223404327696636 Entire dataset(80%) R^2 trained on data (80%): 0.005347569075384029 Holdout R^2 (20%) trained on data (80%): -0.004419322388536706 Dataset D1 R^2 on trained D1: 0.011616915066817723 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004941221409161978 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1758418474699015 | D1-D2 diff: 1.7404386644385612 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2880668582475763 Holdout data R^2 trained on entire dataset(80%): 0.21299713558472877 Dataset D1 R^2 on trained D1: 0.2848264410757001 .................................................. Pipeline #2: Score on D2: 0.14309816076902182 | D1-D2 diff: 1.7541520762596587 Pipeline steps: RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25311095467828837 Holdout data R^2 trained on entire dataset(80%): 0.1707365654641938 Dataset D1 R^2 on trained D1: 0.24871448111332228 .................................................. Pipeline #3: Score on D2: 0.14191743024387615 | D1-D2 diff: 1.792815203022169 Pipeline steps: RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2468666595006651 Holdout data R^2 trained on entire dataset(80%): 0.16919295359700914 Dataset D1 R^2 on trained D1: 0.2387135380571156 .................................................. Pipeline #4: Score on D2: 0.1236923617162532 | D1-D2 diff: 1.86727836244775 Pipeline steps: OverDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22060162300541497 Holdout data R^2 trained on entire dataset(80%): 0.16116350633450882 Dataset D1 R^2 on trained D1: 0.20594763337741606 .................................................. Pipeline #5: Score on D2: 0.12218145806356251 | D1-D2 diff: 2.332350004283771 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.158307154878082 Holdout data R^2 trained on entire dataset(80%): 0.15848734129883302 Dataset D1 R^2 on trained D1: 0.1559743303001513 .................................................. Pipeline #6: Score on D2: 0.10627577120303788 | D1-D2 diff: 3.4059835529331135 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11041879564833146 Holdout data R^2 trained on entire dataset(80%): 0.12543500724361611 Dataset D1 R^2 on trained D1: 0.11370647267823919 .................................................. Pipeline #7: Score on D2: 0.05680550141562235 | D1-D2 diff: 4.227154776779509 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05876110197919693 Holdout data R^2 trained on entire dataset(80%): 0.0673614022178124 Dataset D1 R^2 on trained D1: 0.05993739791193553 .................................................. Pipeline #8: Score on D2: 0.046576408542030046 | D1-D2 diff: 4.294841656639889 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.048100147371381685 Holdout data R^2 trained on entire dataset(80%): 0.061022698949598775 Dataset D1 R^2 on trained D1: 0.04951548830798924 .................................................. Pipeline #9: Score on D2: 0.020716194822898215 | D1-D2 diff: 9.490271749826299 Pipeline steps: DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.020676293856887407 Holdout data R^2 trained on entire dataset(80%): 0.025281398612084782 Dataset D1 R^2 on trained D1: 0.020592916872219424 .................................................. ************************************************************************************** Random Seed 25 - 9 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.18423 Best score on D1-D2 diff: 9.49027 Gen 2 - Best score on D2: 0.19011 Best score on D1-D2 diff: 13.48015 Gen 3 - Best score on D2: 0.19011 Best score on D1-D2 diff: 23.79383 Gen 4 - Best score on D2: 0.19011 Best score on D1-D2 diff: 23.79383 Gen 5 - Best score on D2: 0.19011 Best score on D1-D2 diff: 23.79383 Gen 6 - Best score on D2: 0.19011 Best score on D1-D2 diff: 23.79383 Gen 7 - Best score on D2: 0.19011 Best score on D1-D2 diff: 23.79383 Gen 8 - Best score on D2: 0.19011 Best score on D1-D2 diff: 23.79383 Gen 9 - Best score on D2: 0.19011 Best score on D1-D2 diff: 23.79383 Gen 10 - Best score on D2: 0.19011 Best score on D1-D2 diff: 23.79383 Gen 11 - Best score on D2: 0.19240 Best score on D1-D2 diff: 23.79383 Gen 12 - Best score on D2: 0.19240 Best score on D1-D2 diff: 23.79383 Gen 13 - Best score on D2: 0.19240 Best score on D1-D2 diff: 23.79383 Gen 14 - Best score on D2: 0.19240 Best score on D1-D2 diff: 23.79383 Gen 15 - Best score on D2: 0.19240 Best score on D1-D2 diff: 23.79383 Gen 16 - Best score on D2: 0.19240 Best score on D1-D2 diff: 23.79383 Gen 17 - Best score on D2: 0.19240 Best score on D1-D2 diff: 23.79383 Gen 18 - Best score on D2: 0.19240 Best score on D1-D2 diff: 23.79383 Gen 19 - Best score on D2: 0.19240 Best score on D1-D2 diff: 23.79383 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.00481647401407348 Entire dataset(80%) R^2 trained on data (80%): 0.005514596186097731 Holdout R^2 (20%) trained on data (80%): -0.006947522735856637 Dataset D1 R^2 on trained D1: 0.010306764305270666 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004491589730921897 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.192404489460691 | D1-D2 diff: 1.4323237877059385 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=2, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4219398087161528 Holdout data R^2 trained on entire dataset(80%): 0.24962230743076375 Dataset D1 R^2 on trained D1: 0.4299983320514046 .................................................. Pipeline #2: Score on D2: 0.19010606309272338 | D1-D2 diff: 1.4894640086454591 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=7, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.38530211841255735 Holdout data R^2 trained on entire dataset(80%): 0.24376741196803164 Dataset D1 R^2 on trained D1: 0.39328559076515124 .................................................. Pipeline #3: Score on D2: 0.18930599573990836 | D1-D2 diff: 1.5067464229362315 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.4, min_samples_leaf=10, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3781457895449396 Holdout data R^2 trained on entire dataset(80%): 0.2430677983201035 Dataset D1 R^2 on trained D1: 0.38332278959374777 .................................................. Pipeline #4: Score on D2: 0.18908988737500398 | D1-D2 diff: 1.597515640499851 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=8, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3403567694703922 Holdout data R^2 trained on entire dataset(80%): 0.23908185236850688 Dataset D1 R^2 on trained D1: 0.34262917621851485 .................................................. Pipeline #5: Score on D2: 0.18728354403962988 | D1-D2 diff: 1.6438269181420977 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=7, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.32347198665530286 Holdout data R^2 trained on entire dataset(80%): 0.23507671633992622 Dataset D1 R^2 on trained D1: 0.3242378364826316 .................................................. Pipeline #6: Score on D2: 0.18631925635419377 | D1-D2 diff: 1.7009244016943472 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=15, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30852388933710817 Holdout data R^2 trained on entire dataset(80%): 0.23904931719694889 Dataset D1 R^2 on trained D1: 0.30578955613372183 .................................................. Pipeline #7: Score on D2: 0.1861265860867134 | D1-D2 diff: 1.754473610372315 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=14, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29662386027321674 Holdout data R^2 trained on entire dataset(80%): 0.23189167685426437 Dataset D1 R^2 on trained D1: 0.2916655044901556 .................................................. Pipeline #8: Score on D2: 0.18548142782925048 | D1-D2 diff: 1.79998256279384 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28362249515738003 Holdout data R^2 trained on entire dataset(80%): 0.2355998867817305 Dataset D1 R^2 on trained D1: 0.2807449880987869 .................................................. Pipeline #9: Score on D2: 0.18330905516424212 | D1-D2 diff: 1.8146065566317264 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28072955127293564 Holdout data R^2 trained on entire dataset(80%): 0.23157264354439333 Dataset D1 R^2 on trained D1: 0.27553860671399444 .................................................. Pipeline #10: Score on D2: 0.179446480684874 | D1-D2 diff: 1.8157366785771705 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2766138271709859 Holdout data R^2 trained on entire dataset(80%): 0.23284325521554572 Dataset D1 R^2 on trained D1: 0.27144663031427374 .................................................. Pipeline #11: Score on D2: 0.1793758196013857 | D1-D2 diff: 1.8240251609609097 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27667952448889754 Holdout data R^2 trained on entire dataset(80%): 0.22795831913038256 Dataset D1 R^2 on trained D1: 0.26971511540218907 .................................................. Pipeline #12: Score on D2: 0.17901813597970428 | D1-D2 diff: 1.884606720603583 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=15, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2667981254318921 Holdout data R^2 trained on entire dataset(80%): 0.23340719465303505 Dataset D1 R^2 on trained D1: 0.25828963267023375 .................................................. Pipeline #13: Score on D2: 0.17541368957912606 | D1-D2 diff: 1.9208161985451597 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25865342306551076 Holdout data R^2 trained on entire dataset(80%): 0.2298544256273144 Dataset D1 R^2 on trained D1: 0.24887467704386323 .................................................. Pipeline #14: Score on D2: 0.17375881846158736 | D1-D2 diff: 1.9288269919705436 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=15, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2555289335303187 Holdout data R^2 trained on entire dataset(80%): 0.22442778058685198 Dataset D1 R^2 on trained D1: 0.24600699663794456 .................................................. Pipeline #15: Score on D2: 0.17183186993512434 | D1-D2 diff: 1.938400383522441 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), VarianceThreshold(threshold=0.15), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25240825066372663 Holdout data R^2 trained on entire dataset(80%): 0.2227477069476188 Dataset D1 R^2 on trained D1: 0.24266330678691495 .................................................. Pipeline #16: Score on D2: 0.17052425711387997 | D1-D2 diff: 1.9636381799623825 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=14, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24581570969414523 Holdout data R^2 trained on entire dataset(80%): 0.22157123127579492 Dataset D1 R^2 on trained D1: 0.23778383372571066 .................................................. Pipeline #17: Score on D2: 0.164558918871327 | D1-D2 diff: 1.9963439731063957 Pipeline steps: SelectPercentile(percentile=85), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), FeatureEncodingFrequencySelector(threshold=0.1), VarianceThreshold(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23339552137685027 Holdout data R^2 trained on entire dataset(80%): 0.2128843657927565 Dataset D1 R^2 on trained D1: 0.22751801841393993 .................................................. Pipeline #18: Score on D2: 0.1572155237214229 | D1-D2 diff: 2.1121766586652404 Pipeline steps: VarianceThreshold(threshold=0.1), HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=15, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21938945460149883 Holdout data R^2 trained on entire dataset(80%): 0.20600143553206607 Dataset D1 R^2 on trained D1: 0.20745892642269093 .................................................. Pipeline #19: Score on D2: 0.15648511201017457 | D1-D2 diff: 2.120606581759513 Pipeline steps: VarianceThreshold(threshold=0.1), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=15, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21662557202916233 Holdout data R^2 trained on entire dataset(80%): 0.2033060669204454 Dataset D1 R^2 on trained D1: 0.20593434732098626 .................................................. Pipeline #20: Score on D2: 0.14105511437449858 | D1-D2 diff: 2.179373468620976 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20042251189648597 Holdout data R^2 trained on entire dataset(80%): 0.18996375217097128 Dataset D1 R^2 on trained D1: 0.1853826269599015 .................................................. Pipeline #21: Score on D2: 0.13624567736816895 | D1-D2 diff: 2.241083941318023 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=15, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1919992064887016 Holdout data R^2 trained on entire dataset(80%): 0.1841672346162453 Dataset D1 R^2 on trained D1: 0.17588876801046882 .................................................. Pipeline #22: Score on D2: 0.13417516760286907 | D1-D2 diff: 2.286669724818254 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.2), VarianceThreshold(threshold=0.15), VarianceThreshold(), RandomForestRegressor(max_features=0.1, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18894356507508603 Holdout data R^2 trained on entire dataset(80%): 0.1821950470967787 Dataset D1 R^2 on trained D1: 0.1707503274456419 .................................................. Pipeline #23: Score on D2: 0.12764107069378905 | D1-D2 diff: 3.287547610982343 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=70), VarianceThreshold(threshold=0.1), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09329953521950873 Holdout data R^2 trained on entire dataset(80%): 0.11233676860723729 Dataset D1 R^2 on trained D1: 0.1362018203217027 .................................................. Pipeline #24: Score on D2: 0.1059161752290465 | D1-D2 diff: 3.3042183009847124 Pipeline steps: SelectPercentile(percentile=70), VarianceThreshold(threshold=0.1), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11304955736353162 Holdout data R^2 trained on entire dataset(80%): 0.12663652399562753 Dataset D1 R^2 on trained D1: 0.11430546259335117 .................................................. Pipeline #25: Score on D2: 0.09943433492487685 | D1-D2 diff: 5.822023479995792 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=85), DecisionTreeRegressor(max_depth=2, min_samples_leaf=10, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08481883463373951 Holdout data R^2 trained on entire dataset(80%): 0.10992671899682571 Dataset D1 R^2 on trained D1: 0.09856396437115023 .................................................. Pipeline #26: Score on D2: 0.09943433492487674 | D1-D2 diff: 5.822023479996163 Pipeline steps: SelectPercentile(percentile=85), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08481883463373951 Holdout data R^2 trained on entire dataset(80%): 0.10992671899682571 Dataset D1 R^2 on trained D1: 0.09856396437115034 .................................................. Pipeline #27: Score on D2: 0.060083405601367046 | D1-D2 diff: 6.52450012025522 Pipeline steps: SelectPercentile(percentile=55), VarianceThreshold(threshold=0.35), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09271450396249403 Holdout data R^2 trained on entire dataset(80%): 0.11756931801735171 Dataset D1 R^2 on trained D1: 0.05953156833441975 .................................................. Pipeline #28: Score on D2: 0.020716194813390487 | D1-D2 diff: 9.490270903434881 Pipeline steps: DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.02067629382504288 Holdout data R^2 trained on entire dataset(80%): 0.025281398605198735 Dataset D1 R^2 on trained D1: 0.02059291681873343 .................................................. Pipeline #29: Score on D2: 0.020188891752679017 | D1-D2 diff: 23.793834397199817 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=5), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.02023725062188897 Holdout data R^2 trained on entire dataset(80%): 0.019594594905418372 Dataset D1 R^2 on trained D1: 0.020185771841275013 .................................................. ************************************************************************************** Random Seed 26 - 0 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.00066 Best score on D1-D2 diff: 5.24676 Gen 2 - Best score on D2: 0.00066 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00066 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00066 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.00098 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.00830147302453521 Entire dataset(80%) R^2 trained on data (80%): 0.0039817693200157045 Holdout R^2 (20%) trained on data (80%): -0.008702749131231702 Dataset D1 R^2 on trained D1: 0.008110196794470448 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0022362205000361346 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.0009826481405738052 | D1-D2 diff: 6.967894665887231 Pipeline steps: SelectPercentile(percentile=50), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0012286576490219137 Holdout data R^2 trained on entire dataset(80%): 0.000446838979552 Dataset D1 R^2 on trained D1: 0.001406870634559132 .................................................. Pipeline #2: Score on D2: 0.000797173781456384 | D1-D2 diff: 12.998527508953067 Pipeline steps: SelectPercentile(percentile=25), RecessiveEncoder(), OverDominanceEncoder(), VarianceThreshold(threshold=0.2), RandomForestRegressor(max_features=0.05, min_samples_leaf=9, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002966812084228887 Holdout data R^2 trained on entire dataset(80%): -0.0029159364364934603 Dataset D1 R^2 on trained D1: 0.0008322024290000618 .................................................. Pipeline #3: Score on D2: -3.2015594250811574e-05 | D1-D2 diff: 13.480151627526027 Pipeline steps: VarianceThreshold(threshold=0.25), HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021467281316e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516763899763 Dataset D1 R^2 on trained D1: -1.73106056000627e-06 .................................................. ************************************************************************************** Random Seed 26 - 1 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.03818 Best score on D1-D2 diff: 5.24676 Gen 2 - Best score on D2: 0.05214 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.05214 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.05214 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.05711 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.05711 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.05854 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.05854 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.06199 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.06199 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.06199 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.06199 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.06199 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.005001973984397168 Entire dataset(80%) R^2 trained on data (80%): 0.0036851522604323117 Holdout R^2 (20%) trained on data (80%): -0.0063394206764593175 Dataset D1 R^2 on trained D1: 0.006209415062042867 Combined Dataset (100%) R^2 trained on combined data (100%): 0.002357973807298963 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.06199294984019066 | D1-D2 diff: 1.6724701759514602 Pipeline steps: VarianceThreshold(threshold=0.2), HeterosisEncoder(), RandomForestRegressor(max_features=0.8500000000000001, min_samples_leaf=18, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1864509434677626 Holdout data R^2 trained on entire dataset(80%): 0.06742515377370795 Dataset D1 R^2 on trained D1: 0.18980343161600544 .................................................. Pipeline #2: Score on D2: 0.056608216111212806 | D1-D2 diff: 1.7011646015708948 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), RandomForestRegressor(max_features=0.9000000000000001, min_samples_leaf=20, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17581936119658925 Holdout data R^2 trained on entire dataset(80%): 0.0662297533927324 Dataset D1 R^2 on trained D1: 0.17601105463755617 .................................................. Pipeline #3: Score on D2: 0.054260818464001215 | D1-D2 diff: 1.737620351481347 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1625460206851953 Holdout data R^2 trained on entire dataset(80%): 0.06313303658798575 Dataset D1 R^2 on trained D1: 0.1639541992900807 .................................................. Pipeline #4: Score on D2: 0.04455400780637675 | D1-D2 diff: 1.7668121009685378 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.45, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15304063193820538 Holdout data R^2 trained on entire dataset(80%): 0.05633581098834317 Dataset D1 R^2 on trained D1: 0.14717555063265997 .................................................. Pipeline #5: Score on D2: 0.043538921026737065 | D1-D2 diff: 1.8125052034469857 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13629674807460201 Holdout data R^2 trained on entire dataset(80%): 0.04373474760210594 Dataset D1 R^2 on trained D1: 0.1361969274269701 .................................................. Pipeline #6: Score on D2: 0.03976047330628274 | D1-D2 diff: 1.9836420141310465 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11149648304884163 Holdout data R^2 trained on entire dataset(80%): 0.04940959939178535 Dataset D1 R^2 on trained D1: 0.10434772533881742 .................................................. Pipeline #7: Score on D2: 0.039190849378790915 | D1-D2 diff: 2.0575486532704037 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09698770300706472 Holdout data R^2 trained on entire dataset(80%): 0.045144482857215285 Dataset D1 R^2 on trained D1: 0.09498639723406344 .................................................. Pipeline #8: Score on D2: 0.02770559041717302 | D1-D2 diff: 2.1273303088980726 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), RandomForestRegressor(max_features=0.1, min_samples_leaf=16, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08220713330268081 Holdout data R^2 trained on entire dataset(80%): 0.030798543864955952 Dataset D1 R^2 on trained D1: 0.076532618272949 .................................................. Pipeline #9: Score on D2: 0.021365507623921087 | D1-D2 diff: 2.1766952341811425 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=95), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07433571496468605 Holdout data R^2 trained on entire dataset(80%): 0.02543369276224805 Dataset D1 R^2 on trained D1: 0.0659115878062102 .................................................. Pipeline #10: Score on D2: 0.021009221877051676 | D1-D2 diff: 2.2400798448590304 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07270140674967707 Holdout data R^2 trained on entire dataset(80%): 0.025385923407049482 Dataset D1 R^2 on trained D1: 0.06072343901732258 .................................................. Pipeline #11: Score on D2: 0.012562543140588689 | D1-D2 diff: 2.311502625887641 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.012557509778325149 Holdout data R^2 trained on entire dataset(80%): -0.02142056989736907 Dataset D1 R^2 on trained D1: 0.047591115345130275 .................................................. Pipeline #12: Score on D2: 0.004167749195993831 | D1-D2 diff: 2.7257673616890647 Pipeline steps: VarianceThreshold(threshold=0.15), DominantEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.023908438125410436 Holdout data R^2 trained on entire dataset(80%): 0.0011919695854015755 Dataset D1 R^2 on trained D1: 0.022283020891564753 .................................................. Pipeline #13: Score on D2: 0.003831326853285577 | D1-D2 diff: 2.8406234191883097 Pipeline steps: UnderDominanceEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.02199624498932584 Holdout data R^2 trained on entire dataset(80%): 0.00022139335364146362 Dataset D1 R^2 on trained D1: 0.019189704673103547 .................................................. Pipeline #14: Score on D2: 0.0035563338411961354 | D1-D2 diff: 2.888307527272329 Pipeline steps: DominantEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01919222229416706 Holdout data R^2 trained on entire dataset(80%): -0.0010867644779994912 Dataset D1 R^2 on trained D1: 0.017925324806528198 .................................................. Pipeline #15: Score on D2: 0.002073731436938453 | D1-D2 diff: 4.324770437248005 Pipeline steps: SelectPercentile(percentile=55), HeterosisEncoder(), SelectPercentile(percentile=60), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0028785560322079906 Holdout data R^2 trained on entire dataset(80%): -0.0023314832482927983 Dataset D1 R^2 on trained D1: 0.004932294392468006 .................................................. Pipeline #16: Score on D2: 0.0018369894681526944 | D1-D2 diff: 5.536377899892221 Pipeline steps: SelectPercentile(percentile=25), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0025963622209401738 Holdout data R^2 trained on entire dataset(80%): -0.0036367151124121477 Dataset D1 R^2 on trained D1: 0.0029013704218039393 .................................................. Pipeline #17: Score on D2: 0.0014324116255818309 | D1-D2 diff: 6.930371989701743 Pipeline steps: SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.1, min_samples_leaf=11, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0012740177835773858 Holdout data R^2 trained on entire dataset(80%): -0.003209983190725074 Dataset D1 R^2 on trained D1: 0.0009989268836340104 .................................................. Pipeline #18: Score on D2: 0.001187775646102729 | D1-D2 diff: 8.358787100118732 Pipeline steps: SelectPercentile(percentile=25), SelectPercentile(percentile=25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0011132067873450557 Holdout data R^2 trained on entire dataset(80%): -0.002428657411855939 Dataset D1 R^2 on trained D1: 0.0009829299023172977 .................................................. Pipeline #19: Score on D2: 0.0007971737862666473 | D1-D2 diff: 12.998528337227706 Pipeline steps: SelectPercentile(percentile=25), RecessiveEncoder(), SelectPercentile(percentile=25), RecessiveEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=15, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0008296814197709113 Holdout data R^2 trained on entire dataset(80%): -0.0018415583014483872 Dataset D1 R^2 on trained D1: 0.0008322024248821336 .................................................. Pipeline #20: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=3, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 26 - 2 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.09976 Best score on D1-D2 diff: 5.24676 Gen 2 - Best score on D2: 0.10175 Best score on D1-D2 diff: 13.48015 Gen 3 - Best score on D2: 0.10211 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.10286 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.10300 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.10445 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.10445 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.10445 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.10445 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.10445 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.10445 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.52783 Gen 25 - Best score on D2: 0.10614 Best score on D1-D2 diff: 13.52783 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0044826710170140505 Entire dataset(80%) R^2 trained on data (80%): 0.00408627178687937 Holdout R^2 (20%) trained on data (80%): -0.003717839783012211 Dataset D1 R^2 on trained D1: 0.006792593824331505 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0030872936148336194 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.10614160545047635 | D1-D2 diff: 1.6920052727753836 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=95), VarianceThreshold(threshold=0.2), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=18, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2084476627188908 Holdout data R^2 trained on entire dataset(80%): 0.10559460667091114 Dataset D1 R^2 on trained D1: 0.2281509661793113 .................................................. Pipeline #2: Score on D2: 0.1045348762166084 | D1-D2 diff: 1.7151483534326006 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), VarianceThreshold(threshold=0.2), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=18, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2036489803906355 Holdout data R^2 trained on entire dataset(80%): 0.10555361379886075 Dataset D1 R^2 on trained D1: 0.22009107227980096 .................................................. Pipeline #3: Score on D2: 0.10432061767663747 | D1-D2 diff: 1.7232066824802468 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), VarianceThreshold(threshold=0.2), HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=18, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20568495334879155 Holdout data R^2 trained on entire dataset(80%): 0.10509084280588665 Dataset D1 R^2 on trained D1: 0.21773040069090277 .................................................. Pipeline #4: Score on D2: 0.10390617952976322 | D1-D2 diff: 1.7466608047967853 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), VarianceThreshold(threshold=0.2), HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=20, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1970457184940564 Holdout data R^2 trained on entire dataset(80%): 0.1058090444491816 Dataset D1 R^2 on trained D1: 0.21134610563898326 .................................................. Pipeline #5: Score on D2: 0.10335728712508097 | D1-D2 diff: 1.7492804457736437 Pipeline steps: VarianceThreshold(threshold=0.05), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18794863913830984 Holdout data R^2 trained on entire dataset(80%): 0.09209585682024946 Dataset D1 R^2 on trained D1: 0.2101550693723574 .................................................. Pipeline #6: Score on D2: 0.0963413757836028 | D1-D2 diff: 2.464779374063197 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=20), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004972900175167494 Holdout data R^2 trained on entire dataset(80%): -0.006876779906799024 Dataset D1 R^2 on trained D1: 0.12343629096822128 .................................................. Pipeline #7: Score on D2: 0.09223234128904145 | D1-D2 diff: 2.477326195035092 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0034993181939375617 Holdout data R^2 trained on entire dataset(80%): 0.00112023145856599 Dataset D1 R^2 on trained D1: 0.11878250605127672 .................................................. Pipeline #8: Score on D2: 0.027537351977853897 | D1-D2 diff: 2.58334979959255 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=60), FeatureEncodingFrequencySelector(threshold=0.05), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03383902396818894 Holdout data R^2 trained on entire dataset(80%): 0.006740687737984197 Dataset D1 R^2 on trained D1: 0.04998997765761326 .................................................. Pipeline #9: Score on D2: 0.019382406725138912 | D1-D2 diff: 2.847860745743416 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=60), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=14, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.026063624394961793 Holdout data R^2 trained on entire dataset(80%): 0.005341592328884204 Dataset D1 R^2 on trained D1: 0.03458525643795174 .................................................. Pipeline #10: Score on D2: 0.017579655219898793 | D1-D2 diff: 2.956774652792135 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=60), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.023343127596921143 Holdout data R^2 trained on entire dataset(80%): 0.006795989704547867 Dataset D1 R^2 on trained D1: 0.030663250271733666 .................................................. Pipeline #11: Score on D2: 0.012950351891196199 | D1-D2 diff: 3.659568756566698 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=20), RecessiveEncoder(), SelectPercentile(percentile=70), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002681956641491068 Holdout data R^2 trained on entire dataset(80%): -0.004714387840951906 Dataset D1 R^2 on trained D1: 0.018525807409881856 .................................................. Pipeline #12: Score on D2: 0.002173849398131922 | D1-D2 diff: 9.71765388298813 Pipeline steps: SelectPercentile(percentile=40), UnderDominanceEncoder(), SelectPercentile(percentile=25), OverDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=3, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.003055080610928851 Holdout data R^2 trained on entire dataset(80%): -0.003865484143124487 Dataset D1 R^2 on trained D1: 0.0022859877822251073 .................................................. Pipeline #13: Score on D2: 0.0018076660316772575 | D1-D2 diff: 10.688105926194156 Pipeline steps: SelectPercentile(percentile=40), UnderDominanceEncoder(), SelectPercentile(percentile=25), DecisionTreeRegressor(max_depth=8, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0028040941687941245 Holdout data R^2 trained on entire dataset(80%): -0.002498594960464784 Dataset D1 R^2 on trained D1: 0.001731036353295412 .................................................. Pipeline #14: Score on D2: 0.0018076360321390705 | D1-D2 diff: 10.689150678974663 Pipeline steps: SelectPercentile(percentile=40), UnderDominanceEncoder(), SelectPercentile(percentile=25), DecisionTreeRegressor(max_depth=2, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0023591474650686006 Holdout data R^2 trained on entire dataset(80%): -0.004699541986474154 Dataset D1 R^2 on trained D1: 0.0017310363083664626 .................................................. Pipeline #15: Score on D2: 0.0007809988005366142 | D1-D2 diff: 11.719634798798218 Pipeline steps: SelectPercentile(percentile=5), DecisionTreeRegressor(max_depth=1, min_samples_leaf=20, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0008314417944335073 Holdout data R^2 trained on entire dataset(80%): -0.0018336786327051158 Dataset D1 R^2 on trained D1: 0.0008340070772285824 .................................................. Pipeline #16: Score on D2: 0.000668495198711816 | D1-D2 diff: 13.527833995100096 Pipeline steps: SelectPercentile(percentile=40), UnderDominanceEncoder(), SelectPercentile(percentile=25), HeterosisEncoder(), DecisionTreeRegressor(max_depth=8, min_samples_leaf=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0008370892067155467 Holdout data R^2 trained on entire dataset(80%): -0.002209289181620777 Dataset D1 R^2 on trained D1: 0.0006983550017738427 .................................................. ************************************************************************************** Random Seed 26 - 3 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.11179 Best score on D1-D2 diff: 6.96789 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0035638562666684415 Entire dataset(80%) R^2 trained on data (80%): 0.003623279281668057 Holdout R^2 (20%) trained on data (80%): -0.005476234870023466 Dataset D1 R^2 on trained D1: 0.005819683428909372 Combined Dataset (100%) R^2 trained on combined data (100%): 0.002484942976995197 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.11179188283849695 | D1-D2 diff: 1.654870939805455 Pipeline steps: UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2547158496390486 Holdout data R^2 trained on entire dataset(80%): 0.10771356310049285 Dataset D1 R^2 on trained D1: 0.24512667289336387 .................................................. Pipeline #2: Score on D2: 0.1074534781375811 | D1-D2 diff: 1.6554240774452924 Pipeline steps: UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25083274151962454 Holdout data R^2 trained on entire dataset(80%): 0.1121032055806459 Dataset D1 R^2 on trained D1: 0.24061014940049363 .................................................. Pipeline #3: Score on D2: 0.0696748211080046 | D1-D2 diff: 1.7064047750985352 Pipeline steps: RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2092642577113053 Holdout data R^2 trained on entire dataset(80%): 0.09012170732938218 Dataset D1 R^2 on trained D1: 0.18761771212873046 .................................................. Pipeline #4: Score on D2: 0.0681715291742011 | D1-D2 diff: 1.709328076170977 Pipeline steps: RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2002834575377911 Holdout data R^2 trained on entire dataset(80%): 0.07968983865783186 Dataset D1 R^2 on trained D1: 0.18530966160151452 .................................................. Pipeline #5: Score on D2: 0.05511900973622397 | D1-D2 diff: 1.8374498314415018 Pipeline steps: OverDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16180536118953748 Holdout data R^2 trained on entire dataset(80%): 0.06269565718195613 Dataset D1 R^2 on trained D1: 0.14284697147448955 .................................................. Pipeline #6: Score on D2: 0.055014216287570195 | D1-D2 diff: 2.239135158897614 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07736809189396754 Holdout data R^2 trained on entire dataset(80%): 0.033474545906707065 Dataset D1 R^2 on trained D1: 0.09479549720061531 .................................................. Pipeline #7: Score on D2: 0.03139356749858946 | D1-D2 diff: 2.3165373531542697 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11418263878771429 Holdout data R^2 trained on entire dataset(80%): 0.08172593896404234 Dataset D1 R^2 on trained D1: 0.06611860875428877 .................................................. Pipeline #8: Score on D2: 0.000982648136471087 | D1-D2 diff: 6.967894616160796 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0012409901945801272 Holdout data R^2 trained on entire dataset(80%): 0.000545633894393438 Dataset D1 R^2 on trained D1: 0.0014068706425662825 .................................................. ************************************************************************************** Random Seed 26 - 4 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.13918 Best score on D1-D2 diff: 5.24676 Gen 2 - Best score on D2: 0.14505 Best score on D1-D2 diff: 5.64207 Gen 3 - Best score on D2: 0.14505 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.14574 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.14574 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.003796275077653233 Entire dataset(80%) R^2 trained on data (80%): 0.0046736602370914815 Holdout R^2 (20%) trained on data (80%): -0.008567544393240034 Dataset D1 R^2 on trained D1: 0.006835824739847118 Combined Dataset (100%) R^2 trained on combined data (100%): 0.002947904547503377 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.14574315013912953 | D1-D2 diff: 1.6419221107909492 Pipeline steps: OverDominanceEncoder(), RandomForestRegressor(max_features=0.7500000000000001, min_samples_leaf=17, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28351420089043033 Holdout data R^2 trained on entire dataset(80%): 0.13589613902679165 Dataset D1 R^2 on trained D1: 0.28333407656453713 .................................................. Pipeline #2: Score on D2: 0.1449169472677302 | D1-D2 diff: 1.7413227847609651 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24890581025456593 Holdout data R^2 trained on entire dataset(80%): 0.13477345519366202 Dataset D1 R^2 on trained D1: 0.2536803707700276 .................................................. Pipeline #3: Score on D2: 0.14364424272029852 | D1-D2 diff: 1.766105860110255 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24428728066571848 Holdout data R^2 trained on entire dataset(80%): 0.1384642738362064 Dataset D1 R^2 on trained D1: 0.24643003167163202 .................................................. Pipeline #4: Score on D2: 0.1207654214325149 | D1-D2 diff: 1.9700515079283634 Pipeline steps: VarianceThreshold(threshold=0.2), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=17, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19770613126336123 Holdout data R^2 trained on entire dataset(80%): 0.12065479494055031 Dataset D1 R^2 on trained D1: 0.18715343520571448 .................................................. Pipeline #5: Score on D2: 0.09274249152146552 | D1-D2 diff: 2.4977726184910476 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=20, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10331476680552265 Holdout data R^2 trained on entire dataset(80%): 0.07546973112592525 Dataset D1 R^2 on trained D1: 0.11843392864265667 .................................................. Pipeline #6: Score on D2: 0.021285484963022072 | D1-D2 diff: 2.7391767869468007 Pipeline steps: DominantEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04729184393612884 Holdout data R^2 trained on entire dataset(80%): 0.01956550744956409 Dataset D1 R^2 on trained D1: 0.03904862540874188 .................................................. Pipeline #7: Score on D2: 0.020566484108549665 | D1-D2 diff: 2.841754944116454 Pipeline steps: DominantEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04637111548345196 Holdout data R^2 trained on entire dataset(80%): 0.0198171023639786 Dataset D1 R^2 on trained D1: 0.03590041504830588 .................................................. Pipeline #8: Score on D2: 0.002060069938823683 | D1-D2 diff: 4.477518452184267 Pipeline steps: DominantEncoder(), VarianceThreshold(threshold=0.05), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004673463745262452 Holdout data R^2 trained on entire dataset(80%): -0.003102638384872014 Dataset D1 R^2 on trained D1: 0.004548070436336316 .................................................. Pipeline #9: Score on D2: 0.001187775646102729 | D1-D2 diff: 8.358787100118732 Pipeline steps: SelectPercentile(percentile=25), SelectPercentile(percentile=25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0011132067873450557 Holdout data R^2 trained on entire dataset(80%): -0.002428657411855939 Dataset D1 R^2 on trained D1: 0.0009829299023172977 .................................................. Pipeline #10: Score on D2: 0.0008126580513190262 | D1-D2 diff: 10.071355090309558 Pipeline steps: SelectPercentile(percentile=10), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=5, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0005701231620338643 Holdout data R^2 trained on entire dataset(80%): -0.0012924439090211681 Dataset D1 R^2 on trained D1: 0.0009098540455507731 .................................................. Pipeline #11: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=5, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 26 - 5 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.14790 Best score on D1-D2 diff: 3.91090 Gen 2 - Best score on D2: 0.14851 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.15052 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15052 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.15052 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.15052 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15052 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.15052 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.15407 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.15633 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.15633 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.15633 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.15633 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.15633 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.15708 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.15708 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.15708 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.15708 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.15708 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.15708 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.15708 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.15708 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.15708 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.15811 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.15811 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.005148252585179813 Entire dataset(80%) R^2 trained on data (80%): 0.005655981804162047 Holdout R^2 (20%) trained on data (80%): -0.004868282928950585 Dataset D1 R^2 on trained D1: 0.009323289099238519 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004190399707643233 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1581136904248457 | D1-D2 diff: 1.4540318011871893 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=2, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.35556013351932025 Holdout data R^2 trained on entire dataset(80%): 0.13274426450733434 Dataset D1 R^2 on trained D1: 0.3818334683757202 .................................................. Pipeline #2: Score on D2: 0.15638496448436456 | D1-D2 diff: 1.484149989696068 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=2, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3562879859756223 Holdout data R^2 trained on entire dataset(80%): 0.1505721447513707 Dataset D1 R^2 on trained D1: 0.36249010625696465 .................................................. Pipeline #3: Score on D2: 0.1540660506607996 | D1-D2 diff: 1.5525195000818945 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=8, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31582706034528885 Holdout data R^2 trained on entire dataset(80%): 0.14792943255043745 Dataset D1 R^2 on trained D1: 0.3261941388158246 .................................................. Pipeline #4: Score on D2: 0.1509829352315717 | D1-D2 diff: 1.6307599815777318 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), VarianceThreshold(threshold=0.2), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2871748810485838 Holdout data R^2 trained on entire dataset(80%): 0.14456341750665858 Dataset D1 R^2 on trained D1: 0.2923798125056515 .................................................. Pipeline #5: Score on D2: 0.14980312346311353 | D1-D2 diff: 1.648970748360458 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=11, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2823513209358355 Holdout data R^2 trained on entire dataset(80%): 0.14588252110762534 Dataset D1 R^2 on trained D1: 0.28505652418375627 .................................................. Pipeline #6: Score on D2: 0.14977607592595343 | D1-D2 diff: 1.7016912002941418 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), VarianceThreshold(threshold=0.2), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26373204229366354 Holdout data R^2 trained on entire dataset(80%): 0.14574181660573815 Dataset D1 R^2 on trained D1: 0.26903118329720654 .................................................. Pipeline #7: Score on D2: 0.14856882317336217 | D1-D2 diff: 1.8257661054811312 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23736930213312213 Holdout data R^2 trained on entire dataset(80%): 0.1429252624095828 Dataset D1 R^2 on trained D1: 0.2385640422799633 .................................................. Pipeline #8: Score on D2: 0.1435961991165693 | D1-D2 diff: 1.8370179357269234 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2322184314455782 Holdout data R^2 trained on entire dataset(80%): 0.1396370461612615 Dataset D1 R^2 on trained D1: 0.2314066917744284 .................................................. Pipeline #9: Score on D2: 0.14275948119331838 | D1-D2 diff: 1.8371848015450138 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RecessiveEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23148715108539752 Holdout data R^2 trained on entire dataset(80%): 0.13818988208192384 Dataset D1 R^2 on trained D1: 0.2305380759745319 .................................................. Pipeline #10: Score on D2: 0.14225205879267266 | D1-D2 diff: 1.8407796375808598 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=19, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23092677619262758 Holdout data R^2 trained on entire dataset(80%): 0.1397047009090664 Dataset D1 R^2 on trained D1: 0.22934697258888426 .................................................. Pipeline #11: Score on D2: 0.1403569054257896 | D1-D2 diff: 1.8830314766085314 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2257435264322254 Holdout data R^2 trained on entire dataset(80%): 0.13995792670235652 Dataset D1 R^2 on trained D1: 0.21989399242629892 .................................................. Pipeline #12: Score on D2: 0.13490977370906287 | D1-D2 diff: 1.9017016965217568 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2119776086238384 Holdout data R^2 trained on entire dataset(80%): 0.13126797597504847 Dataset D1 R^2 on trained D1: 0.21136909248248703 .................................................. Pipeline #13: Score on D2: 0.13339609838312927 | D1-D2 diff: 1.9178946642713472 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21193528464259992 Holdout data R^2 trained on entire dataset(80%): 0.13701529141379853 Dataset D1 R^2 on trained D1: 0.20730572296851968 .................................................. Pipeline #14: Score on D2: 0.11933775825909432 | D1-D2 diff: 1.9500556319206581 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=13, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20084767748619248 Holdout data R^2 trained on entire dataset(80%): 0.12182671386944688 Dataset D1 R^2 on trained D1: 0.18849091263845275 .................................................. Pipeline #15: Score on D2: 0.10924163864768832 | D1-D2 diff: 2.087980150306314 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16864669695185053 Holdout data R^2 trained on entire dataset(80%): 0.11692838202083855 Dataset D1 R^2 on trained D1: 0.16185481689125358 .................................................. Pipeline #16: Score on D2: 0.10364278691578455 | D1-D2 diff: 2.1259707181169247 Pipeline steps: SelectPercentile(percentile=90), HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1775381919593958 Holdout data R^2 trained on entire dataset(80%): 0.12142525785444813 Dataset D1 R^2 on trained D1: 0.15259483716185318 .................................................. Pipeline #17: Score on D2: 0.08743526149881364 | D1-D2 diff: 3.125483888678537 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09311764248784526 Holdout data R^2 trained on entire dataset(80%): 0.08533680002051502 Dataset D1 R^2 on trained D1: 0.09791452936827916 .................................................. Pipeline #18: Score on D2: 0.08611064810029256 | D1-D2 diff: 3.1637709705094306 Pipeline steps: VarianceThreshold(threshold=0.15), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09248711001490184 Holdout data R^2 trained on entire dataset(80%): 0.08330808704209225 Dataset D1 R^2 on trained D1: 0.09609178133128182 .................................................. Pipeline #19: Score on D2: 0.033292889733652076 | D1-D2 diff: 3.4872556930056975 Pipeline steps: DominantEncoder(), RecessiveEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=15, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.030161729476016697 Holdout data R^2 trained on entire dataset(80%): 0.026349524229005827 Dataset D1 R^2 on trained D1: 0.02653105063582295 .................................................. Pipeline #20: Score on D2: 0.027404632340373936 | D1-D2 diff: 3.5999201578141786 Pipeline steps: SelectPercentile(percentile=15), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.029908982834894182 Holdout data R^2 trained on entire dataset(80%): 0.033307604834075244 Dataset D1 R^2 on trained D1: 0.03335890235482264 .................................................. Pipeline #21: Score on D2: 0.027166466780695675 | D1-D2 diff: 3.7462986759914467 Pipeline steps: DominantEncoder(), OverDominanceEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=4, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03262668727933615 Holdout data R^2 trained on entire dataset(80%): 0.021709779301171017 Dataset D1 R^2 on trained D1: 0.03224327087182377 .................................................. Pipeline #22: Score on D2: 0.027117707799278357 | D1-D2 diff: 3.9196147421711354 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=5), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=11, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.029308619614289544 Holdout data R^2 trained on entire dataset(80%): 0.0331805812313547 Dataset D1 R^2 on trained D1: 0.031354395678107494 .................................................. Pipeline #23: Score on D2: 0.01997803042841595 | D1-D2 diff: 4.507656197955757 Pipeline steps: SelectPercentile(percentile=15), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.02109801427787683 Holdout data R^2 trained on entire dataset(80%): 0.026486985558414555 Dataset D1 R^2 on trained D1: 0.022400157151409195 .................................................. Pipeline #24: Score on D2: 0.01959890478982773 | D1-D2 diff: 4.647598751927257 Pipeline steps: SelectPercentile(percentile=15), OverDominanceEncoder(), SelectPercentile(percentile=35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.020727807580666502 Holdout data R^2 trained on entire dataset(80%): 0.025611085125458066 Dataset D1 R^2 on trained D1: 0.0217422172777596 .................................................. Pipeline #25: Score on D2: 0.007719524022607449 | D1-D2 diff: 7.233627684701917 Pipeline steps: SelectPercentile(percentile=35), DominantEncoder(), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.008063105604964815 Holdout data R^2 trained on entire dataset(80%): 0.009104651831663091 Dataset D1 R^2 on trained D1: 0.008084761548868835 .................................................. Pipeline #26: Score on D2: 0.005990723607309345 | D1-D2 diff: 7.707339158541743 Pipeline steps: SelectPercentile(percentile=15), DominantEncoder(), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.006885490390982185 Holdout data R^2 trained on entire dataset(80%): 0.007594734636398259 Dataset D1 R^2 on trained D1: 0.0062741120394878 .................................................. Pipeline #27: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.8, min_samples_leaf=8, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 26 - 6 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.15736 Best score on D1-D2 diff: 3.91090 Gen 2 - Best score on D2: 0.15736 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.15736 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15910 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.15910 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.15910 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15910 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.16310 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.16310 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.16310 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.16310 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.16310 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.16310 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.16310 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.16450 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.006075960215791154 Entire dataset(80%) R^2 trained on data (80%): 0.005378845562818313 Holdout R^2 (20%) trained on data (80%): -0.005685873947697484 Dataset D1 R^2 on trained D1: 0.00970497676525739 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003872227311093157 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1644959806535914 | D1-D2 diff: 1.447975433799049 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=6, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3799527978951227 Holdout data R^2 trained on entire dataset(80%): 0.16529297830015022 Dataset D1 R^2 on trained D1: 0.3919822691290843 .................................................. Pipeline #2: Score on D2: 0.16070264162209258 | D1-D2 diff: 1.5773383403771741 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=6, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31591880518759985 Holdout data R^2 trained on entire dataset(80%): 0.16747640313238887 Dataset D1 R^2 on trained D1: 0.3222502603036198 .................................................. Pipeline #3: Score on D2: 0.16019475003466765 | D1-D2 diff: 1.5796254085922161 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=6, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31561228619769743 Holdout data R^2 trained on entire dataset(80%): 0.16546153302914668 Dataset D1 R^2 on trained D1: 0.3208088086788683 .................................................. Pipeline #4: Score on D2: 0.15752218072468815 | D1-D2 diff: 1.6625668726754845 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2803908434126282 Holdout data R^2 trained on entire dataset(80%): 0.15958893481660408 Dataset D1 R^2 on trained D1: 0.28840526148103973 .................................................. Pipeline #5: Score on D2: 0.1570264836555061 | D1-D2 diff: 1.7223204217282915 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=20, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26739206251191394 Holdout data R^2 trained on entire dataset(80%): 0.15745546562712232 Dataset D1 R^2 on trained D1: 0.2706698776636738 .................................................. Pipeline #6: Score on D2: 0.1543461540674621 | D1-D2 diff: 1.7597927069823693 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(max_features=0.45, min_samples_leaf=17, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2521605873391979 Holdout data R^2 trained on entire dataset(80%): 0.14764065941662108 Dataset D1 R^2 on trained D1: 0.25861485093224135 .................................................. Pipeline #7: Score on D2: 0.15288916367022432 | D1-D2 diff: 1.760174993154927 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), VarianceThreshold(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=17, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25158433426961957 Holdout data R^2 trained on entire dataset(80%): 0.14709850462535146 Dataset D1 R^2 on trained D1: 0.2570673070449425 .................................................. Pipeline #8: Score on D2: 0.15085564828803188 | D1-D2 diff: 1.8221156328072823 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RecessiveEncoder(), RecessiveEncoder(), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2409620834954438 Holdout data R^2 trained on entire dataset(80%): 0.15456626684627317 Dataset D1 R^2 on trained D1: 0.24157423239891662 .................................................. Pipeline #9: Score on D2: 0.14903749465943827 | D1-D2 diff: 1.8228299028479955 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=20, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23607555067910713 Holdout data R^2 trained on entire dataset(80%): 0.15095797798764987 Dataset D1 R^2 on trained D1: 0.23961397118137917 .................................................. Pipeline #10: Score on D2: 0.14767096053388173 | D1-D2 diff: 1.8788298915348276 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2308659401423554 Holdout data R^2 trained on entire dataset(80%): 0.15069554370594163 Dataset D1 R^2 on trained D1: 0.22792190565686876 .................................................. Pipeline #11: Score on D2: 0.13894655345250306 | D1-D2 diff: 1.8794392328345395 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.3, min_samples_leaf=19, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21694627672904876 Holdout data R^2 trained on entire dataset(80%): 0.14494946090972816 Dataset D1 R^2 on trained D1: 0.21909347512369048 .................................................. Pipeline #12: Score on D2: 0.13521025967966027 | D1-D2 diff: 1.8956999668573853 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21586378758588098 Holdout data R^2 trained on entire dataset(80%): 0.14649464897520248 Dataset D1 R^2 on trained D1: 0.21264245815408278 .................................................. Pipeline #13: Score on D2: 0.12636356605653876 | D1-D2 diff: 1.9051742211052778 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2041576245364216 Holdout data R^2 trained on entire dataset(80%): 0.12988621697880376 Dataset D1 R^2 on trained D1: 0.20226696329505145 .................................................. Pipeline #14: Score on D2: 0.1223047072736767 | D1-D2 diff: 2.3987284111997864 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=3, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13317962931260108 Holdout data R^2 trained on entire dataset(80%): 0.11761323824626635 Dataset D1 R^2 on trained D1: 0.15250948775293371 .................................................. Pipeline #15: Score on D2: 0.09501284924934317 | D1-D2 diff: 2.4776224306568246 Pipeline steps: VarianceThreshold(threshold=0.15), SelectPercentile(percentile=60), HeterosisEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=5, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10361658661116413 Holdout data R^2 trained on entire dataset(80%): 0.0842737338423526 Dataset D1 R^2 on trained D1: 0.12155031846281372 .................................................. Pipeline #16: Score on D2: 0.08751223358847582 | D1-D2 diff: 3.2652799518544513 Pipeline steps: VarianceThreshold(threshold=0.15), SelectPercentile(percentile=65), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09305369146131048 Holdout data R^2 trained on entire dataset(80%): 0.08190880481441987 Dataset D1 R^2 on trained D1: 0.0963089038419751 .................................................. Pipeline #17: Score on D2: 0.033292889733652076 | D1-D2 diff: 3.4872556930056975 Pipeline steps: UnderDominanceEncoder(), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03016172947601692 Holdout data R^2 trained on entire dataset(80%): 0.026349524229005827 Dataset D1 R^2 on trained D1: 0.02653105063582295 .................................................. Pipeline #18: Score on D2: 0.02863463690874013 | D1-D2 diff: 3.751804113299013 Pipeline steps: SelectPercentile(percentile=45), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03979692255294487 Holdout data R^2 trained on entire dataset(80%): 0.04404341648232435 Dataset D1 R^2 on trained D1: 0.03368170750129873 .................................................. Pipeline #19: Score on D2: 0.027480548735686727 | D1-D2 diff: 3.810689369528154 Pipeline steps: OverDominanceEncoder(), VarianceThreshold(), SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.7500000000000001, min_samples_leaf=10, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.029946075793518157 Holdout data R^2 trained on entire dataset(80%): 0.03300777263582244 Dataset D1 R^2 on trained D1: 0.03222281359666623 .................................................. Pipeline #20: Score on D2: 0.027117707799278357 | D1-D2 diff: 3.9196147421711354 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.7500000000000001, min_samples_leaf=2, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.029308619614289544 Holdout data R^2 trained on entire dataset(80%): 0.0331805812313547 Dataset D1 R^2 on trained D1: 0.031354395678107494 .................................................. Pipeline #21: Score on D2: 0.00806076544464418 | D1-D2 diff: 9.688964248603828 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=10), RandomForestRegressor(max_features=0.8, min_samples_leaf=19, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.008129288200587337 Holdout data R^2 trained on entire dataset(80%): 0.016996147398551265 Dataset D1 R^2 on trained D1: 0.008174237935015594 .................................................. Pipeline #22: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: DominantEncoder(), DominantEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.7000000000000001, min_samples_leaf=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 26 - 7 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.16247 Best score on D1-D2 diff: 3.45922 Gen 2 - Best score on D2: 0.16247 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.16334 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.16405 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.16572 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.17224 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.17252 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.17252 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.17252 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.17252 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.00826911745538883 Entire dataset(80%) R^2 trained on data (80%): 0.006694581106454711 Holdout R^2 (20%) trained on data (80%): -0.003125615018839012 Dataset D1 R^2 on trained D1: 0.013829542046587262 Combined Dataset (100%) R^2 trained on combined data (100%): 0.005356676550949624 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.17251892333183483 | D1-D2 diff: 1.4954853155925003 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=3, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3694230795867359 Holdout data R^2 trained on entire dataset(80%): 0.17322905875637806 Dataset D1 R^2 on trained D1: 0.3724458951727039 .................................................. Pipeline #2: Score on D2: 0.16739814726496205 | D1-D2 diff: 1.5723727774003131 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=8, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3264130835592881 Holdout data R^2 trained on entire dataset(80%): 0.16793603369839294 Dataset D1 R^2 on trained D1: 0.33099612652804034 .................................................. Pipeline #3: Score on D2: 0.16391110476156645 | D1-D2 diff: 1.6131448081910633 Pipeline steps: VarianceThreshold(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=15, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30872987793515627 Holdout data R^2 trained on entire dataset(80%): 0.17021432238606649 Dataset D1 R^2 on trained D1: 0.31158596927974946 .................................................. Pipeline #4: Score on D2: 0.16255725723300163 | D1-D2 diff: 1.652190925218729 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2942832916289472 Holdout data R^2 trained on entire dataset(80%): 0.16873230733375089 Dataset D1 R^2 on trained D1: 0.29675928260756834 .................................................. Pipeline #5: Score on D2: 0.16082104810391418 | D1-D2 diff: 1.6527747090178786 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29535659069497555 Holdout data R^2 trained on entire dataset(80%): 0.17044435580830364 Dataset D1 R^2 on trained D1: 0.29483356557103646 .................................................. Pipeline #6: Score on D2: 0.15940551003763892 | D1-D2 diff: 1.7169660443139978 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27184567342243393 Holdout data R^2 trained on entire dataset(80%): 0.16493153861890464 Dataset D1 R^2 on trained D1: 0.2744731416858106 .................................................. Pipeline #7: Score on D2: 0.1556825923106986 | D1-D2 diff: 1.719266451522961 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=19, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26749876393847605 Holdout data R^2 trained on entire dataset(80%): 0.16380777009128789 Dataset D1 R^2 on trained D1: 0.2701356092352478 .................................................. Pipeline #8: Score on D2: 0.15412287157480875 | D1-D2 diff: 1.7527671725281757 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2584620869168168 Holdout data R^2 trained on entire dataset(80%): 0.1610429689281211 Dataset D1 R^2 on trained D1: 0.26007338777836886 .................................................. Pipeline #9: Score on D2: 0.153559556699158 | D1-D2 diff: 1.7925151404950663 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25622497902494734 Holdout data R^2 trained on entire dataset(80%): 0.16304834295101844 Dataset D1 R^2 on trained D1: 0.25042049448887627 .................................................. Pipeline #10: Score on D2: 0.15313869906939004 | D1-D2 diff: 1.801615572788608 Pipeline steps: VarianceThreshold(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2538257292771362 Holdout data R^2 trained on entire dataset(80%): 0.1656314324142063 Dataset D1 R^2 on trained D1: 0.2480573356718332 .................................................. Pipeline #11: Score on D2: 0.15280085618441042 | D1-D2 diff: 1.813820791003912 Pipeline steps: VarianceThreshold(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24985367658928914 Holdout data R^2 trained on entire dataset(80%): 0.1652713994086903 Dataset D1 R^2 on trained D1: 0.24519033073885665 .................................................. Pipeline #12: Score on D2: 0.15276000966962655 | D1-D2 diff: 1.822541266059588 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2478548956400637 Holdout data R^2 trained on entire dataset(80%): 0.1637939148773 Dataset D1 R^2 on trained D1: 0.2433938783944628 .................................................. Pipeline #13: Score on D2: 0.15116214430354213 | D1-D2 diff: 1.8248162833849704 Pipeline steps: VarianceThreshold(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2427207052142758 Holdout data R^2 trained on entire dataset(80%): 0.16555423514653234 Dataset D1 R^2 on trained D1: 0.2413448808277423 .................................................. Pipeline #14: Score on D2: 0.1457737778054512 | D1-D2 diff: 1.8296264396843167 Pipeline steps: VarianceThreshold(threshold=0.2), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23600702488999092 Holdout data R^2 trained on entire dataset(80%): 0.15560476913167476 Dataset D1 R^2 on trained D1: 0.23501187260847423 .................................................. Pipeline #15: Score on D2: 0.14558218634388798 | D1-D2 diff: 1.8354747628074055 Pipeline steps: VarianceThreshold(threshold=0.2), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23553769259283563 Holdout data R^2 trained on entire dataset(80%): 0.15138514050283736 Dataset D1 R^2 on trained D1: 0.23368835783880426 .................................................. Pipeline #16: Score on D2: 0.1391675161679936 | D1-D2 diff: 1.9049401013433294 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22356443855173735 Holdout data R^2 trained on entire dataset(80%): 0.15462679914461674 Dataset D1 R^2 on trained D1: 0.21510823481430097 .................................................. Pipeline #17: Score on D2: 0.1264907325220469 | D1-D2 diff: 2.032146390852779 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=19, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1929108577489682 Holdout data R^2 trained on entire dataset(80%): 0.1387285895733027 Dataset D1 R^2 on trained D1: 0.1851288526680438 .................................................. Pipeline #18: Score on D2: 0.11586853155783894 | D1-D2 diff: 2.2288180658589267 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=10, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14470105307491732 Holdout data R^2 trained on entire dataset(80%): 0.13864546705567948 Dataset D1 R^2 on trained D1: 0.1563915253259084 .................................................. Pipeline #19: Score on D2: 0.09031358882569374 | D1-D2 diff: 2.2796120386006793 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(threshold=0.25), DecisionTreeRegressor(max_depth=3, min_samples_leaf=18, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11114153503285906 Holdout data R^2 trained on entire dataset(80%): 0.11550666092118911 Dataset D1 R^2 on trained D1: 0.12734380375360388 .................................................. Pipeline #20: Score on D2: 0.08867275050125056 | D1-D2 diff: 2.531782464821554 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10250102807968164 Holdout data R^2 trained on entire dataset(80%): 0.09714163568361689 Dataset D1 R^2 on trained D1: 0.11301128640978586 .................................................. Pipeline #21: Score on D2: 0.08681797885621789 | D1-D2 diff: 3.0687087438445575 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.092690652224798 Holdout data R^2 trained on entire dataset(80%): 0.08531458784856483 Dataset D1 R^2 on trained D1: 0.09809455655958799 .................................................. Pipeline #22: Score on D2: 0.030880481780452662 | D1-D2 diff: 3.6577874062005113 Pipeline steps: UnderDominanceEncoder(), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=9, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.034142278265503756 Holdout data R^2 trained on entire dataset(80%): 0.01268240295393619 Dataset D1 R^2 on trained D1: 0.03646680627124965 .................................................. Pipeline #23: Score on D2: 0.02425851711217375 | D1-D2 diff: 4.068700802450049 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=25), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.026152498894151055 Holdout data R^2 trained on entire dataset(80%): 0.034850446699669635 Dataset D1 R^2 on trained D1: 0.02790754331273082 .................................................. Pipeline #24: Score on D2: 0.00499354800180718 | D1-D2 diff: 7.726854963691636 Pipeline steps: DominantEncoder(), VarianceThreshold(threshold=0.2), DecisionTreeRegressor(max_depth=1, min_samples_leaf=6, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004890070150986969 Holdout data R^2 trained on entire dataset(80%): -0.0005307313466929031 Dataset D1 R^2 on trained D1: 0.004713011770601128 .................................................. Pipeline #25: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=3, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 26 - 8 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.16490 Best score on D1-D2 diff: 4.85089 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0007529611029808425 Entire dataset(80%) R^2 trained on data (80%): 0.008456703519512887 Holdout R^2 (20%) trained on data (80%): -0.011044102641605491 Dataset D1 R^2 on trained D1: 0.010853327485486841 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0058015854532349476 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.16490243389774073 | D1-D2 diff: 1.641731284308978 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29497265754744995 Holdout data R^2 trained on entire dataset(80%): 0.18000019638534692 Dataset D1 R^2 on trained D1: 0.30255734295134307 .................................................. Pipeline #2: Score on D2: 0.15593736241742073 | D1-D2 diff: 1.643507148841596 Pipeline steps: UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2921177364073262 Holdout data R^2 trained on entire dataset(80%): 0.16379241796610222 Dataset D1 R^2 on trained D1: 0.2929982721297224 .................................................. Pipeline #3: Score on D2: 0.13259031451418413 | D1-D2 diff: 1.684214189240989 Pipeline steps: RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25469938144866533 Holdout data R^2 trained on entire dataset(80%): 0.14119884365662105 Dataset D1 R^2 on trained D1: 0.25687302390814915 .................................................. Pipeline #4: Score on D2: 0.13110879078142545 | D1-D2 diff: 1.6993950382320044 Pipeline steps: RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24584819143263947 Holdout data R^2 trained on entire dataset(80%): 0.14048766484245245 Dataset D1 R^2 on trained D1: 0.2510097386530252 .................................................. Pipeline #5: Score on D2: 0.13099334398367934 | D1-D2 diff: 1.7229644732261313 Pipeline steps: RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2445890724167462 Holdout data R^2 trained on entire dataset(80%): 0.1399852092747199 Dataset D1 R^2 on trained D1: 0.24446691169500145 .................................................. Pipeline #6: Score on D2: 0.11926411958644434 | D1-D2 diff: 1.8162637387991138 Pipeline steps: OverDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22379408240088627 Holdout data R^2 trained on entire dataset(80%): 0.13436561842275796 Dataset D1 R^2 on trained D1: 0.21115752587107461 .................................................. Pipeline #7: Score on D2: 0.1067520931366891 | D1-D2 diff: 1.957261423183034 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14024782791434698 Holdout data R^2 trained on entire dataset(80%): 0.13757868999436196 Dataset D1 R^2 on trained D1: 0.17489248931882373 .................................................. Pipeline #8: Score on D2: 0.08693042605629497 | D1-D2 diff: 3.0881852757768247 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09276469293104173 Holdout data R^2 trained on entire dataset(80%): 0.08648207396619301 Dataset D1 R^2 on trained D1: 0.09792520768671475 .................................................. Pipeline #9: Score on D2: 0.024524804872014694 | D1-D2 diff: 3.2738841048347087 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.029111115365682227 Holdout data R^2 trained on entire dataset(80%): 0.03298459124375164 Dataset D1 R^2 on trained D1: 0.03322936429747192 .................................................. Pipeline #10: Score on D2: 0.009654098980489856 | D1-D2 diff: 4.201085165170709 Pipeline steps: DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01136112164983949 Holdout data R^2 trained on entire dataset(80%): 0.011702430116213969 Dataset D1 R^2 on trained D1: 0.012864461357086854 .................................................. Pipeline #11: Score on D2: 0.004564061523262608 | D1-D2 diff: 4.850887546520865 Pipeline steps: VarianceThreshold(threshold=0.35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.007020734674535678 Holdout data R^2 trained on entire dataset(80%): -0.010095524035774739 Dataset D1 R^2 on trained D1: 0.006370050820989737 .................................................. ************************************************************************************** Random Seed 26 - 9 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.16427 Best score on D1-D2 diff: 5.14135 Gen 2 - Best score on D2: 0.16427 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.16427 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.16427 Best score on D1-D2 diff: 11.96036 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0013498830664357975 Entire dataset(80%) R^2 trained on data (80%): 0.007508100299581932 Holdout R^2 (20%) trained on data (80%): -0.009063958497728786 Dataset D1 R^2 on trained D1: 0.010588551617080388 Combined Dataset (100%) R^2 trained on combined data (100%): 0.005381985472886375 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.16426798799639386 | D1-D2 diff: 1.6201952972390192 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2965245154944267 Holdout data R^2 trained on entire dataset(80%): 0.18893011908219493 Dataset D1 R^2 on trained D1: 0.3093890777497068 .................................................. Pipeline #2: Score on D2: 0.15732272384371715 | D1-D2 diff: 1.656972126252921 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=17, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2776347437125586 Holdout data R^2 trained on entire dataset(80%): 0.1895340335546828 Dataset D1 R^2 on trained D1: 0.28998247828127277 .................................................. Pipeline #3: Score on D2: 0.1564912900126917 | D1-D2 diff: 1.6667349370748141 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2749592883690911 Holdout data R^2 trained on entire dataset(80%): 0.19186676967883576 Dataset D1 R^2 on trained D1: 0.2860700573593289 .................................................. Pipeline #4: Score on D2: 0.13048704657258947 | D1-D2 diff: 1.7762255119200476 Pipeline steps: OverDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23229074846381903 Holdout data R^2 trained on entire dataset(80%): 0.14783962470821432 Dataset D1 R^2 on trained D1: 0.23095038038238613 .................................................. Pipeline #5: Score on D2: 0.10240061899049058 | D1-D2 diff: 2.087565401712035 Pipeline steps: OverDominanceEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16632951192767487 Holdout data R^2 trained on entire dataset(80%): 0.12601032584509497 Dataset D1 R^2 on trained D1: 0.1550556215440151 .................................................. Pipeline #6: Score on D2: 0.08706333257993693 | D1-D2 diff: 2.353659490608071 Pipeline steps: SelectPercentile(percentile=50), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11103110185734921 Holdout data R^2 trained on entire dataset(80%): 0.09650953513833394 Dataset D1 R^2 on trained D1: 0.11964891361774443 .................................................. Pipeline #7: Score on D2: 0.032310415422324046 | D1-D2 diff: 2.805397768927601 Pipeline steps: SelectPercentile(percentile=40), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.040883886582589235 Holdout data R^2 trained on entire dataset(80%): 0.045091356825403106 Dataset D1 R^2 on trained D1: 0.04845482663573908 .................................................. Pipeline #8: Score on D2: 0.02297432697103552 | D1-D2 diff: 2.8223913047013203 Pipeline steps: RecessiveEncoder(), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.031480958371359 Holdout data R^2 trained on entire dataset(80%): 0.028497119382831615 Dataset D1 R^2 on trained D1: 0.038733415604769506 .................................................. Pipeline #9: Score on D2: 0.015298620440450605 | D1-D2 diff: 2.9951165081535884 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.021974478483676285 Holdout data R^2 trained on entire dataset(80%): 0.027669101816731434 Dataset D1 R^2 on trained D1: 0.027725014357771305 .................................................. Pipeline #10: Score on D2: 0.012288592148479993 | D1-D2 diff: 3.1354256388508954 Pipeline steps: SelectPercentile(percentile=25), HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=8, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01787111817282727 Holdout data R^2 trained on entire dataset(80%): 0.02315630203130148 Dataset D1 R^2 on trained D1: 0.02263558094278717 .................................................. Pipeline #11: Score on D2: 0.00965409896622449 | D1-D2 diff: 4.201085155890297 Pipeline steps: SelectPercentile(percentile=55), DecisionTreeRegressor(max_depth=1, min_samples_leaf=6, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00866489641234991 Holdout data R^2 trained on entire dataset(80%): 0.0031503730558315457 Dataset D1 R^2 on trained D1: 0.012864461371188907 .................................................. Pipeline #12: Score on D2: 0.004541361949741529 | D1-D2 diff: 5.141345851746344 Pipeline steps: VarianceThreshold(threshold=0.35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.006396190845443406 Holdout data R^2 trained on entire dataset(80%): -0.009392738865733108 Dataset D1 R^2 on trained D1: 0.005972536889007896 .................................................. Pipeline #13: Score on D2: -4.886781085700065e-05 | D1-D2 diff: 11.960360745435652 Pipeline steps: DominantEncoder(), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0 Holdout data R^2 trained on entire dataset(80%): -0.0012028184018280097 Dataset D1 R^2 on trained D1: 0.0 .................................................. ************************************************************************************** Random Seed 27 - 0 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.00291 Best score on D1-D2 diff: 10.57332 Gen 2 - Best score on D2: 0.00291 Best score on D1-D2 diff: 10.57332 Gen 3 - Best score on D2: 0.00291 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.00291 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.00291 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.00291 Best score on D1-D2 diff: 11.96036 Gen 7 - Best score on D2: 0.00444 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0026779306509816525 Entire dataset(80%) R^2 trained on data (80%): 0.005181944909279368 Holdout R^2 (20%) trained on data (80%): -0.008190157838054546 Dataset D1 R^2 on trained D1: 0.004572056749066689 Combined Dataset (100%) R^2 trained on combined data (100%): 0.00394653771382536 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.004440668373968948 | D1-D2 diff: 4.437567537708676 Pipeline steps: SelectPercentile(percentile=45), VarianceThreshold(threshold=0.05), SelectPercentile(percentile=55), VarianceThreshold(threshold=0.35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0038781446919315954 Holdout data R^2 trained on entire dataset(80%): -0.006895873447969603 Dataset D1 R^2 on trained D1: 0.0018618539153095748 .................................................. Pipeline #2: Score on D2: 0.0029091934381378914 | D1-D2 diff: 10.57331823718607 Pipeline steps: SelectPercentile(percentile=25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004196043038482422 Holdout data R^2 trained on entire dataset(80%): -0.007940291789303 Dataset D1 R^2 on trained D1: 0.0029892053751825465 .................................................. Pipeline #3: Score on D2: 0.0026100867218137314 | D1-D2 diff: 11.326017499222354 Pipeline steps: VarianceThreshold(threshold=0.2), VarianceThreshold(threshold=0.35), SelectPercentile(percentile=35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004048327432877086 Holdout data R^2 trained on entire dataset(80%): -0.007287337332390509 Dataset D1 R^2 on trained D1: 0.002670856980746761 .................................................. Pipeline #4: Score on D2: -3.2015594250811574e-05 | D1-D2 diff: 13.480151627526027 Pipeline steps: VarianceThreshold(threshold=0.1), RecessiveEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=8, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021467281316e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516763899763 Dataset D1 R^2 on trained D1: -1.73106056000627e-06 .................................................. ************************************************************************************** Random Seed 27 - 1 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.03916 Best score on D1-D2 diff: 10.57332 Gen 2 - Best score on D2: 0.04056 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.04399 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.04916 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.05209 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.05214 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.05972 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.05972 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.05972 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.08207 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.08207 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.08207 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.08207 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.08207 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.08207 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.08675 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.08675 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.08675 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.08675 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.08675 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.08675 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.08675 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.10103 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.10107 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.10107 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0017040739804798921 Entire dataset(80%) R^2 trained on data (80%): 0.0058406499494614295 Holdout R^2 (20%) trained on data (80%): -0.007109982678521254 Dataset D1 R^2 on trained D1: 0.00489303410569264 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004642985651758469 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.10107261565851522 | D1-D2 diff: 2.2959535533077235 Pipeline steps: SelectPercentile(percentile=70), FeatureEncodingFrequencySelector(threshold=0.1), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.8500000000000001, min_samples_leaf=18, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15258559014949646 Holdout data R^2 trained on entire dataset(80%): 0.08538031643644584 Dataset D1 R^2 on trained D1: 0.13705977837140482 .................................................. Pipeline #2: Score on D2: 0.09941008383514105 | D1-D2 diff: 2.4837857871832494 Pipeline steps: SelectPercentile(percentile=70), FeatureEncodingFrequencySelector(threshold=0.1), HeterosisEncoder(), RandomForestRegressor(max_features=0.8500000000000001, min_samples_leaf=18, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1288809486339494 Holdout data R^2 trained on entire dataset(80%): 0.07220012985070057 Dataset D1 R^2 on trained D1: 0.12568512768772766 .................................................. Pipeline #3: Score on D2: 0.04658252291936227 | D1-D2 diff: 6.491955524710475 Pipeline steps: SelectPercentile(percentile=75), FeatureEncodingFrequencySelector(threshold=0.1), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04912443618158646 Holdout data R^2 trained on entire dataset(80%): 0.041885494425754066 Dataset D1 R^2 on trained D1: 0.047145509258020524 .................................................. Pipeline #4: Score on D2: 0.007932452223974984 | D1-D2 diff: 6.937975955251646 Pipeline steps: SelectPercentile(percentile=35), FeatureEncodingFrequencySelector(threshold=0.1), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.6000000000000001, min_samples_leaf=8, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01148664339082317 Holdout data R^2 trained on entire dataset(80%): -0.01054371633039941 Dataset D1 R^2 on trained D1: 0.007500864743191449 .................................................. Pipeline #5: Score on D2: 0.007766685065749401 | D1-D2 diff: 7.384558777527191 Pipeline steps: SelectPercentile(percentile=35), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=8, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01137560304071239 Holdout data R^2 trained on entire dataset(80%): -0.009666073271128894 Dataset D1 R^2 on trained D1: 0.00743040448130694 .................................................. Pipeline #6: Score on D2: 0.00775324639181052 | D1-D2 diff: 7.480732354609848 Pipeline steps: SelectPercentile(percentile=35), UnderDominanceEncoder(), VarianceThreshold(threshold=0.25), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=8, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011376098827093784 Holdout data R^2 trained on entire dataset(80%): -0.009668225307820189 Dataset D1 R^2 on trained D1: 0.007433928294654835 .................................................. Pipeline #7: Score on D2: 0.007738289264347276 | D1-D2 diff: 8.933272001064582 Pipeline steps: SelectPercentile(percentile=35), FeatureEncodingFrequencySelector(threshold=0.1), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.6000000000000001, min_samples_leaf=8, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01148664339082317 Holdout data R^2 trained on entire dataset(80%): -0.01054371633039941 Dataset D1 R^2 on trained D1: 0.007581268253275986 .................................................. Pipeline #8: Score on D2: 0.007608104043711372 | D1-D2 diff: 9.570771686782853 Pipeline steps: SelectPercentile(percentile=35), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=4, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011437274347502524 Holdout data R^2 trained on entire dataset(80%): -0.010280352692888517 Dataset D1 R^2 on trained D1: 0.007488921630330103 .................................................. Pipeline #9: Score on D2: 0.007607172534159656 | D1-D2 diff: 9.581472460507841 Pipeline steps: SelectPercentile(percentile=35), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=4, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011437039859108378 Holdout data R^2 trained on entire dataset(80%): -0.010265203266865086 Dataset D1 R^2 on trained D1: 0.00748852165041114 .................................................. Pipeline #10: Score on D2: 0.007486035335698471 | D1-D2 diff: 13.25047335354368 Pipeline steps: SelectPercentile(percentile=35), FeatureEncodingFrequencySelector(threshold=0.1), OverDominanceEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011437226800253386 Holdout data R^2 trained on entire dataset(80%): -0.01028430341094122 Dataset D1 R^2 on trained D1: 0.007518474856315294 .................................................. Pipeline #11: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=5, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 27 - 2 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.08171 Best score on D1-D2 diff: 4.10396 Gen 2 - Best score on D2: 0.08464 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.08533 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.08826 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.08826 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.08826 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.08826 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.08826 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.08826 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.08826 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.08826 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.08835 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.08835 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.09095 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.09095 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.09095 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.09297 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.09297 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.09297 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.09297 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.09297 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.09297 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.09297 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.09297 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.09297 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.004613633449569665 Entire dataset(80%) R^2 trained on data (80%): 0.005142803075918967 Holdout R^2 (20%) trained on data (80%): -0.008082704313716604 Dataset D1 R^2 on trained D1: 0.005389986528814172 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003885608239168503 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.09297124632323539 | D1-D2 diff: 2.6714421193226148 Pipeline steps: SelectPercentile(percentile=55), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=20, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0028239590789539415 Holdout data R^2 trained on entire dataset(80%): 0.0025197423253269235 Dataset D1 R^2 on trained D1: 0.11260561420747883 .................................................. Pipeline #2: Score on D2: 0.027705695189569335 | D1-D2 diff: 2.943302970874513 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10149786707710573 Holdout data R^2 trained on entire dataset(80%): 0.09371628056491488 Dataset D1 R^2 on trained D1: 0.04103047756793565 .................................................. Pipeline #3: Score on D2: 0.021569410339218886 | D1-D2 diff: 4.1015318845444595 Pipeline steps: VarianceThreshold(threshold=0.35), DecisionTreeRegressor(max_depth=2, min_samples_leaf=9, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.003626856825373226 Holdout data R^2 trained on entire dataset(80%): -0.00464396309235382 Dataset D1 R^2 on trained D1: 0.02510299606514299 .................................................. Pipeline #4: Score on D2: 0.0015964471510290235 | D1-D2 diff: 5.178311669537252 Pipeline steps: SelectPercentile(percentile=25), VarianceThreshold(threshold=0.35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.003878144687906482 Holdout data R^2 trained on entire dataset(80%): -0.006895873439360711 Dataset D1 R^2 on trained D1: 0.002987191346774609 .................................................. Pipeline #5: Score on D2: 0.0011972127619144235 | D1-D2 diff: 5.222602429721148 Pipeline steps: SelectPercentile(percentile=50), VarianceThreshold(threshold=0.35), DominantEncoder(), RecessiveEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0026728357705222416 Holdout data R^2 trained on entire dataset(80%): -0.005569554211933392 Dataset D1 R^2 on trained D1: 0.0025413763732579087 .................................................. Pipeline #6: Score on D2: 0.0009795223243419526 | D1-D2 diff: 5.277282661797588 Pipeline steps: SelectPercentile(percentile=30), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0026789656505336046 Holdout data R^2 trained on entire dataset(80%): -0.005298680800426947 Dataset D1 R^2 on trained D1: 0.002268835955058335 .................................................. Pipeline #7: Score on D2: 0.0009772605518542088 | D1-D2 diff: 5.701753716606534 Pipeline steps: SelectPercentile(percentile=25), VarianceThreshold(threshold=0.35), RecessiveEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0031592339998263164 Holdout data R^2 trained on entire dataset(80%): -0.005461839545205072 Dataset D1 R^2 on trained D1: 0.0019234240352634657 .................................................. Pipeline #8: Score on D2: 0.0004127258024738678 | D1-D2 diff: 5.815981550548267 Pipeline steps: SelectPercentile(percentile=25), VarianceThreshold(threshold=0.35), VarianceThreshold(threshold=0.15), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0010479625195434217 Holdout data R^2 trained on entire dataset(80%): -0.0010843923898724572 Dataset D1 R^2 on trained D1: 0.0012867187319826057 .................................................. Pipeline #9: Score on D2: 8.992580233357916e-05 | D1-D2 diff: 11.707558798700449 Pipeline steps: SelectPercentile(percentile=5), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=9, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 1.2956388625973148e-05 Holdout data R^2 trained on entire dataset(80%): -0.0009642608163880073 Dataset D1 R^2 on trained D1: 0.000143153123509987 .................................................. Pipeline #10: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: OverDominanceEncoder(), RecessiveEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=18, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 27 - 3 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.10426 Best score on D1-D2 diff: 5.60614 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.005253248544512035 Entire dataset(80%) R^2 trained on data (80%): 0.0052602941017223515 Holdout R^2 (20%) trained on data (80%): -0.003655670272487921 Dataset D1 R^2 on trained D1: 0.007541099642774385 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004447301155285133 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.10426129458118716 | D1-D2 diff: 1.6508562287087403 Pipeline steps: UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2409192260656393 Holdout data R^2 trained on entire dataset(80%): 0.128489056084658 Dataset D1 R^2 on trained D1: 0.2388978491704653 .................................................. Pipeline #2: Score on D2: 0.10238805747698387 | D1-D2 diff: 1.6727914884432602 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23976147857579766 Holdout data R^2 trained on entire dataset(80%): 0.13398462535076594 Dataset D1 R^2 on trained D1: 0.23010036736748563 .................................................. Pipeline #3: Score on D2: 0.08267667245529853 | D1-D2 diff: 1.6970158891919662 Pipeline steps: RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20397738567528245 Holdout data R^2 trained on entire dataset(80%): 0.10641717382268678 Dataset D1 R^2 on trained D1: 0.20325142114922956 .................................................. Pipeline #4: Score on D2: 0.07460385900253674 | D1-D2 diff: 2.5332951685602536 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08774604167984568 Holdout data R^2 trained on entire dataset(80%): 0.09247475622537815 Dataset D1 R^2 on trained D1: 0.09888431398768194 .................................................. Pipeline #5: Score on D2: 0.004613925675685526 | D1-D2 diff: 5.606139313880308 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004135714463174067 Holdout data R^2 trained on entire dataset(80%): 0.0008118064320238316 Dataset D1 R^2 on trained D1: 0.003601543578659827 .................................................. ************************************************************************************** Random Seed 27 - 4 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.10426 Best score on D1-D2 diff: 5.60614 Gen 1 - Best score on D2: 0.11842 Best score on D1-D2 diff: 5.60614 Gen 2 - Best score on D2: 0.12073 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.12284 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.12434 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.12434 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.12434 Best score on D1-D2 diff: 11.96036 Gen 7 - Best score on D2: 0.12896 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.12896 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.13292 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.13292 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.13292 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.005251555871599534 Entire dataset(80%) R^2 trained on data (80%): 0.004420724940950693 Holdout R^2 (20%) trained on data (80%): -0.003021733133217186 Dataset D1 R^2 on trained D1: 0.006226547679229477 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003947759432784248 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.13291700255890893 | D1-D2 diff: 1.5517005685966958 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.2), RandomForestRegressor(max_features=0.4, min_samples_leaf=6, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30014914580084995 Holdout data R^2 trained on entire dataset(80%): 0.15327358374444167 Dataset D1 R^2 on trained D1: 0.305408750415478 .................................................. Pipeline #2: Score on D2: 0.13177807580417067 | D1-D2 diff: 1.8907956830743604 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.4, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2121795436034557 Holdout data R^2 trained on entire dataset(80%): 0.14976061452718847 Dataset D1 R^2 on trained D1: 0.21001676967676186 .................................................. Pipeline #3: Score on D2: 0.09581918299999159 | D1-D2 diff: 1.9508035820805063 Pipeline steps: VarianceThreshold(threshold=0.05), HeterosisEncoder(), OverDominanceEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=16, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18027981622807454 Holdout data R^2 trained on entire dataset(80%): 0.12274370045393435 Dataset D1 R^2 on trained D1: 0.1648663433684565 .................................................. Pipeline #4: Score on D2: 0.09519248630747412 | D1-D2 diff: 1.9623221707365304 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=16, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18025724503730034 Holdout data R^2 trained on entire dataset(80%): 0.1268755740189944 Dataset D1 R^2 on trained D1: 0.16263267200803422 .................................................. Pipeline #5: Score on D2: 0.08092851400629386 | D1-D2 diff: 2.077959248843813 Pipeline steps: VarianceThreshold(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.35), HeterosisEncoder(), SelectPercentile(percentile=65), RandomForestRegressor(max_features=0.2, min_samples_leaf=16, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13826241489212598 Holdout data R^2 trained on entire dataset(80%): 0.09660144782792479 Dataset D1 R^2 on trained D1: 0.1345639598237589 .................................................. Pipeline #6: Score on D2: 0.07460385900253685 | D1-D2 diff: 2.5332951685602567 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(threshold=0.35), DecisionTreeRegressor(max_depth=2, min_samples_leaf=9, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08774604167984557 Holdout data R^2 trained on entire dataset(80%): 0.09247475622537815 Dataset D1 R^2 on trained D1: 0.09888431398768194 .................................................. Pipeline #7: Score on D2: 0.040061233781326866 | D1-D2 diff: 2.559560892386145 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=17, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08774604167984557 Holdout data R^2 trained on entire dataset(80%): 0.09247475622537815 Dataset D1 R^2 on trained D1: 0.0633602796411794 .................................................. Pipeline #8: Score on D2: 0.037121812471846316 | D1-D2 diff: 3.63682465396141 Pipeline steps: SelectPercentile(percentile=95), VarianceThreshold(threshold=0.35), DecisionTreeRegressor(max_depth=2, min_samples_leaf=9, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04062193505508538 Holdout data R^2 trained on entire dataset(80%): 0.032639642749018094 Dataset D1 R^2 on trained D1: 0.04283805372893856 .................................................. Pipeline #9: Score on D2: 0.004613925675685748 | D1-D2 diff: 5.606139313880001 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004135714463174289 Holdout data R^2 trained on entire dataset(80%): 0.0008118064320238316 Dataset D1 R^2 on trained D1: 0.003601543578659827 .................................................. Pipeline #10: Score on D2: 0.004613925675685526 | D1-D2 diff: 5.606139313880308 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004135714463174067 Holdout data R^2 trained on entire dataset(80%): 0.0008118064320238316 Dataset D1 R^2 on trained D1: 0.003601543578659827 .................................................. Pipeline #11: Score on D2: 0.003922823446718793 | D1-D2 diff: 6.783083893804428 Pipeline steps: SelectPercentile(percentile=10), OverDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004184845774201862 Holdout data R^2 trained on entire dataset(80%): -0.000596247632858038 Dataset D1 R^2 on trained D1: 0.004395203168492179 .................................................. Pipeline #12: Score on D2: 0.0006169977529615345 | D1-D2 diff: 7.686866833869491 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), SelectPercentile(percentile=85), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), OverDominanceEncoder(), RecessiveEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0010946799277236074 Holdout data R^2 trained on entire dataset(80%): -0.003783452572901247 Dataset D1 R^2 on trained D1: 0.0009034172449235989 .................................................. Pipeline #13: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: RecessiveEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=2, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 27 - 5 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.12866 Best score on D1-D2 diff: 7.01408 Gen 2 - Best score on D2: 0.14145 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14145 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.14145 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.14303 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.14863 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.14863 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.14863 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.14863 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.14863 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.14863 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.14863 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.14863 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.14863 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.14923 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.011559158108564782 Entire dataset(80%) R^2 trained on data (80%): 0.005122432324231485 Holdout R^2 (20%) trained on data (80%): -0.01023152936374716 Dataset D1 R^2 on trained D1: 0.010766306663278824 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003151787035067577 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.14923231267077297 | D1-D2 diff: 1.500019662741703 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=2, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.35527565833953034 Holdout data R^2 trained on entire dataset(80%): 0.1633477018863012 Dataset D1 R^2 on trained D1: 0.3467528198787545 .................................................. Pipeline #2: Score on D2: 0.14862703617136164 | D1-D2 diff: 1.5506030275207296 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3310166722107921 Holdout data R^2 trained on entire dataset(80%): 0.16394013440172117 Dataset D1 R^2 on trained D1: 0.32160767221428055 .................................................. Pipeline #3: Score on D2: 0.1450947913165178 | D1-D2 diff: 1.6375923152489333 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.25, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2913523208329205 Holdout data R^2 trained on entire dataset(80%): 0.16231761422981905 Dataset D1 R^2 on trained D1: 0.28414666122498056 .................................................. Pipeline #4: Score on D2: 0.13848518906425 | D1-D2 diff: 1.6915960500547673 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), DominantEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=12, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2676969609530202 Holdout data R^2 trained on entire dataset(80%): 0.15406054028737237 Dataset D1 R^2 on trained D1: 0.2606126562957154 .................................................. Pipeline #5: Score on D2: 0.1374461702063544 | D1-D2 diff: 1.7638947123482709 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=14, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2485220491049478 Holdout data R^2 trained on entire dataset(80%): 0.15325313671351193 Dataset D1 R^2 on trained D1: 0.24074832167367055 .................................................. Pipeline #6: Score on D2: 0.1357825176338474 | D1-D2 diff: 1.8016312341016825 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=16, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23970544149859374 Holdout data R^2 trained on entire dataset(80%): 0.15177262408968006 Dataset D1 R^2 on trained D1: 0.23069785382481267 .................................................. Pipeline #7: Score on D2: 0.1347686879872294 | D1-D2 diff: 1.8202532347856566 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), DominantEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=17, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23421558401510456 Holdout data R^2 trained on entire dataset(80%): 0.1506802440623911 Dataset D1 R^2 on trained D1: 0.22585911835373707 .................................................. Pipeline #8: Score on D2: 0.13353816975988897 | D1-D2 diff: 1.8451949078536207 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=18, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2310376822169793 Holdout data R^2 trained on entire dataset(80%): 0.15129842624167966 Dataset D1 R^2 on trained D1: 0.21980245143806965 .................................................. Pipeline #9: Score on D2: 0.13257447390013544 | D1-D2 diff: 1.8746129205688387 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2227914251356844 Holdout data R^2 trained on entire dataset(80%): 0.15235727508390573 Dataset D1 R^2 on trained D1: 0.2135499622567335 .................................................. Pipeline #10: Score on D2: 0.12784786227495126 | D1-D2 diff: 1.8967436393771326 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21450241556844785 Holdout data R^2 trained on entire dataset(80%): 0.14496407033863778 Dataset D1 R^2 on trained D1: 0.20510977483529746 .................................................. Pipeline #11: Score on D2: 0.12355239372935867 | D1-D2 diff: 1.897220571287168 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.25, min_samples_leaf=17, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20980828425363796 Holdout data R^2 trained on entire dataset(80%): 0.1444029812562877 Dataset D1 R^2 on trained D1: 0.2007366457801424 .................................................. Pipeline #12: Score on D2: 0.1163833468520693 | D1-D2 diff: 2.048963431012483 Pipeline steps: VarianceThreshold(threshold=0.1), VarianceThreshold(threshold=0.35), SelectPercentile(percentile=90), HeterosisEncoder(), DecisionTreeRegressor(max_depth=5, min_samples_leaf=6, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1559901352991775 Holdout data R^2 trained on entire dataset(80%): 0.12846625610553997 Dataset D1 R^2 on trained D1: 0.17311992905981555 .................................................. Pipeline #13: Score on D2: 0.105979409506336 | D1-D2 diff: 2.1191899922703747 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), DominantEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=17, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1656389706280048 Holdout data R^2 trained on entire dataset(80%): 0.1153324630658572 Dataset D1 R^2 on trained D1: 0.15556099639683119 .................................................. Pipeline #14: Score on D2: 0.10366826676703667 | D1-D2 diff: 2.1351500515544046 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=18, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16060325639687678 Holdout data R^2 trained on entire dataset(80%): 0.11591316928673334 Dataset D1 R^2 on trained D1: 0.15178392095726578 .................................................. Pipeline #15: Score on D2: 0.10171691420780027 | D1-D2 diff: 2.1386466134839877 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=18, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1616492231139811 Holdout data R^2 trained on entire dataset(80%): 0.11809735645576791 Dataset D1 R^2 on trained D1: 0.14951867414197495 .................................................. Pipeline #16: Score on D2: 0.09882714876648957 | D1-D2 diff: 3.120732813572916 Pipeline steps: VarianceThreshold(threshold=0.1), VarianceThreshold(threshold=0.35), SelectPercentile(percentile=90), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=6, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10336640946182629 Holdout data R^2 trained on entire dataset(80%): 0.10923353564949811 Dataset D1 R^2 on trained D1: 0.1093703780248032 .................................................. Pipeline #17: Score on D2: 0.08545292708579777 | D1-D2 diff: 3.42243861855634 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=55), RecessiveEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.040839113880802524 Holdout data R^2 trained on entire dataset(80%): 0.03271483281944021 Dataset D1 R^2 on trained D1: 0.09274174882254538 .................................................. Pipeline #18: Score on D2: 0.038578556999481095 | D1-D2 diff: 4.07165971791482 Pipeline steps: SelectPercentile(percentile=10), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.005507511313887492 Holdout data R^2 trained on entire dataset(80%): -0.000111424753782563 Dataset D1 R^2 on trained D1: 0.042216987622902935 .................................................. Pipeline #19: Score on D2: 0.03132760231373388 | D1-D2 diff: 4.121267313269116 Pipeline steps: UnderDominanceEncoder(), OverDominanceEncoder(), SelectPercentile(percentile=90), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.031214753101961423 Holdout data R^2 trained on entire dataset(80%): 0.033152076855141654 Dataset D1 R^2 on trained D1: 0.027861216796122523 .................................................. Pipeline #20: Score on D2: 0.03083815664847467 | D1-D2 diff: 4.382179349675383 Pipeline steps: SelectPercentile(percentile=55), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.030816283379459075 Holdout data R^2 trained on entire dataset(80%): 0.03496571497256573 Dataset D1 R^2 on trained D1: 0.02812647059133122 .................................................. Pipeline #21: Score on D2: 0.029981201511036337 | D1-D2 diff: 4.463382247616694 Pipeline steps: SelectPercentile(percentile=35), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.030115398054088027 Holdout data R^2 trained on entire dataset(80%): 0.03412313397889044 Dataset D1 R^2 on trained D1: 0.027461531465769462 .................................................. Pipeline #22: Score on D2: 0.029337434662981554 | D1-D2 diff: 8.268996579304606 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03129249337738549 Holdout data R^2 trained on entire dataset(80%): 0.032552070441399406 Dataset D1 R^2 on trained D1: 0.02912354551478924 .................................................. Pipeline #23: Score on D2: 0.029125530116009046 | D1-D2 diff: 9.19023736788332 Pipeline steps: UnderDominanceEncoder(), OverDominanceEncoder(), SelectPercentile(percentile=90), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03161526801630454 Holdout data R^2 trained on entire dataset(80%): 0.03338589436739703 Dataset D1 R^2 on trained D1: 0.028985347846892995 .................................................. Pipeline #24: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: DominantEncoder(), OverDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=1.0, min_samples_leaf=17, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 27 - 6 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.14254 Best score on D1-D2 diff: 6.12620 Gen 2 - Best score on D2: 0.14335 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14426 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.15085 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.15396 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.15396 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15905 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.15905 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.15905 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.15905 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.15905 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.15905 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.15905 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.15905 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.15905 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.009517507979570627 Entire dataset(80%) R^2 trained on data (80%): 0.0063714087113661 Holdout R^2 (20%) trained on data (80%): -0.011601244914884346 Dataset D1 R^2 on trained D1: 0.011386798397251208 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0038443943251150747 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.15904728774099952 | D1-D2 diff: 1.480618003060441 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=2, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3746080036188423 Holdout data R^2 trained on entire dataset(80%): 0.18460710852089868 Dataset D1 R^2 on trained D1: 0.36712611767459 .................................................. Pipeline #2: Score on D2: 0.1567948514833395 | D1-D2 diff: 1.5671615666510734 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33199401743911383 Holdout data R^2 trained on entire dataset(80%): 0.18705391071417476 Dataset D1 R^2 on trained D1: 0.3225797280955114 .................................................. Pipeline #3: Score on D2: 0.15576818952722526 | D1-D2 diff: 1.5776041612311638 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=2, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3283137374430375 Holdout data R^2 trained on entire dataset(80%): 0.1804146404186303 Dataset D1 R^2 on trained D1: 0.31720695486092165 .................................................. Pipeline #4: Score on D2: 0.15328937285684907 | D1-D2 diff: 1.6240161116956773 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31001000345841545 Holdout data R^2 trained on entire dataset(80%): 0.17801940853598353 Dataset D1 R^2 on trained D1: 0.2970495720747416 .................................................. Pipeline #5: Score on D2: 0.1509543489755698 | D1-D2 diff: 1.6421560233543457 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=5, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2964770174798167 Holdout data R^2 trained on entire dataset(80%): 0.17616230521032383 Dataset D1 R^2 on trained D1: 0.28846689704984363 .................................................. Pipeline #6: Score on D2: 0.14848126507074866 | D1-D2 diff: 1.6501427186855733 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), RandomForestRegressor(max_features=0.3, min_samples_leaf=9, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28940226999493146 Holdout data R^2 trained on entire dataset(80%): 0.17556227469998964 Dataset D1 R^2 on trained D1: 0.28335083430655983 .................................................. Pipeline #7: Score on D2: 0.14481760655477127 | D1-D2 diff: 1.724914472776565 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=16, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2697408698786474 Holdout data R^2 trained on entire dataset(80%): 0.1776989798032953 Dataset D1 R^2 on trained D1: 0.2577789205376455 .................................................. Pipeline #8: Score on D2: 0.143172038639648 | D1-D2 diff: 1.7291830143140037 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=16, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26852205926047845 Holdout data R^2 trained on entire dataset(80%): 0.17521268234003728 Dataset D1 R^2 on trained D1: 0.255022081973197 .................................................. Pipeline #9: Score on D2: 0.14070948228975744 | D1-D2 diff: 1.7468822873472043 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=17, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2599943857251755 Holdout data R^2 trained on entire dataset(80%): 0.17513087814646133 Dataset D1 R^2 on trained D1: 0.24809493067292554 .................................................. Pipeline #10: Score on D2: 0.14064753744369374 | D1-D2 diff: 1.7921273038299723 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=16, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.248743870972293 Holdout data R^2 trained on entire dataset(80%): 0.175941241395272 Dataset D1 R^2 on trained D1: 0.23759234967559972 .................................................. Pipeline #11: Score on D2: 0.13954671078325198 | D1-D2 diff: 1.7943223618783406 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24778323756034892 Holdout data R^2 trained on entire dataset(80%): 0.17422158697897971 Dataset D1 R^2 on trained D1: 0.23601800872252932 .................................................. Pipeline #12: Score on D2: 0.13937113257801825 | D1-D2 diff: 1.8178309812814912 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.25, min_samples_leaf=13, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23963749057710015 Holdout data R^2 trained on entire dataset(80%): 0.16440978376055382 Dataset D1 R^2 on trained D1: 0.23094804495435484 .................................................. Pipeline #13: Score on D2: 0.13790998764413454 | D1-D2 diff: 1.8298191951434253 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=19, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2405370523799255 Holdout data R^2 trained on entire dataset(80%): 0.17095799817896884 Dataset D1 R^2 on trained D1: 0.22711048657498134 .................................................. Pipeline #14: Score on D2: 0.13737640242077098 | D1-D2 diff: 1.862920313245952 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23004170244119482 Holdout data R^2 trained on entire dataset(80%): 0.16702502075965275 Dataset D1 R^2 on trained D1: 0.2204040793814802 .................................................. Pipeline #15: Score on D2: 0.1350209814208826 | D1-D2 diff: 1.9097658033186702 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=17, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21757157193544485 Holdout data R^2 trained on entire dataset(80%): 0.16296519304214774 Dataset D1 R^2 on trained D1: 0.21019703960661562 .................................................. Pipeline #16: Score on D2: 0.13216929664159183 | D1-D2 diff: 1.9686872867442076 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21117944558268364 Holdout data R^2 trained on entire dataset(80%): 0.15873197943722928 Dataset D1 R^2 on trained D1: 0.19874151868144896 .................................................. Pipeline #17: Score on D2: 0.1257829203482287 | D1-D2 diff: 2.000274969626315 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20436940299776396 Holdout data R^2 trained on entire dataset(80%): 0.15222137045046935 Dataset D1 R^2 on trained D1: 0.1882485609554878 .................................................. Pipeline #18: Score on D2: 0.11947058788190446 | D1-D2 diff: 2.5892282802130393 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=5, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1299039385790064 Holdout data R^2 trained on entire dataset(80%): 0.11963314975234829 Dataset D1 R^2 on trained D1: 0.14172000472488067 .................................................. Pipeline #19: Score on D2: 0.08545292708579777 | D1-D2 diff: 3.42243861855634 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08921921766549779 Holdout data R^2 trained on entire dataset(80%): 0.08658527638577795 Dataset D1 R^2 on trained D1: 0.09274174882254538 .................................................. Pipeline #20: Score on D2: 0.038578556999481095 | D1-D2 diff: 4.07165971791482 Pipeline steps: SelectPercentile(percentile=10), HeterosisEncoder(), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.018674664383863004 Holdout data R^2 trained on entire dataset(80%): 0.01046522511123038 Dataset D1 R^2 on trained D1: 0.042216987622902935 .................................................. Pipeline #21: Score on D2: 0.03478660869737449 | D1-D2 diff: 6.126199234047604 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03564446757308948 Holdout data R^2 trained on entire dataset(80%): 0.04026833620763115 Dataset D1 R^2 on trained D1: 0.03549657139893747 .................................................. Pipeline #22: Score on D2: 0.008823334820506012 | D1-D2 diff: 9.250424154969217 Pipeline steps: VarianceThreshold(), RecessiveEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.010927249130136207 Holdout data R^2 trained on entire dataset(80%): 0.017604393104563854 Dataset D1 R^2 on trained D1: 0.00868676541676705 .................................................. Pipeline #23: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=9, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 27 - 7 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.15416 Best score on D1-D2 diff: 4.97796 Gen 2 - Best score on D2: 0.15923 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.16469 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.16469 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.16469 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.16469 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.16845 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.16845 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.16845 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.16845 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.16845 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.01137359082089695 Entire dataset(80%) R^2 trained on data (80%): 0.007311637659183301 Holdout R^2 (20%) trained on data (80%): -0.011197379352541503 Dataset D1 R^2 on trained D1: 0.013348716443677966 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004707781734224148 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1684513779312694 | D1-D2 diff: 1.51716359678636 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=4, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3627509058736723 Holdout data R^2 trained on entire dataset(80%): 0.21327606229574403 Dataset D1 R^2 on trained D1: 0.3571941571361734 .................................................. Pipeline #2: Score on D2: 0.16807209723733596 | D1-D2 diff: 1.5522605017584434 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=4, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34363373439480205 Holdout data R^2 trained on entire dataset(80%): 0.20703551644638918 Dataset D1 R^2 on trained D1: 0.34031509405641724 .................................................. Pipeline #3: Score on D2: 0.1605123909500954 | D1-D2 diff: 1.5654749541603346 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=10, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3275009937116301 Holdout data R^2 trained on entire dataset(80%): 0.20582647527022568 Dataset D1 R^2 on trained D1: 0.3270128767006353 .................................................. Pipeline #4: Score on D2: 0.1569117298081213 | D1-D2 diff: 1.6284287388641627 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(max_features=0.8, min_samples_leaf=13, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3047578800903923 Holdout data R^2 trained on entire dataset(80%): 0.199426141809519 Dataset D1 R^2 on trained D1: 0.29912003702414214 .................................................. Pipeline #5: Score on D2: 0.15662412717552188 | D1-D2 diff: 1.6764630828441285 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=16, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2865265330028074 Holdout data R^2 trained on entire dataset(80%): 0.19712567354711086 Dataset D1 R^2 on trained D1: 0.2832213045382711 .................................................. Pipeline #6: Score on D2: 0.1541569241310733 | D1-D2 diff: 1.6907814170257065 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2777505617503647 Holdout data R^2 trained on entire dataset(80%): 0.19322545721471895 Dataset D1 R^2 on trained D1: 0.27651992977735984 .................................................. Pipeline #7: Score on D2: 0.15239910681449165 | D1-D2 diff: 1.7684799669941942 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2680186146600424 Holdout data R^2 trained on entire dataset(80%): 0.1945110438390112 Dataset D1 R^2 on trained D1: 0.25463406453200477 .................................................. Pipeline #8: Score on D2: 0.15206105669700887 | D1-D2 diff: 1.7925564038245796 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26053261380146175 Holdout data R^2 trained on entire dataset(80%): 0.19169978637163754 Dataset D1 R^2 on trained D1: 0.24891307612459068 .................................................. Pipeline #9: Score on D2: 0.15006059043781694 | D1-D2 diff: 2.030662134512931 Pipeline steps: VarianceThreshold(threshold=0.25), SelectPercentile(percentile=60), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16023180789804503 Holdout data R^2 trained on entire dataset(80%): 0.1169057930181121 Dataset D1 R^2 on trained D1: 0.20887033828938695 .................................................. Pipeline #10: Score on D2: 0.1278403921070892 | D1-D2 diff: 2.113313618869738 Pipeline steps: VarianceThreshold(threshold=0.25), SelectPercentile(percentile=60), UnderDominanceEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12534436283771577 Holdout data R^2 trained on entire dataset(80%): 0.08857738558883366 Dataset D1 R^2 on trained D1: 0.1779757584698567 .................................................. Pipeline #11: Score on D2: 0.12316316158387608 | D1-D2 diff: 2.1222570198308075 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=19, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18140250950582382 Holdout data R^2 trained on entire dataset(80%): 0.1477052130790415 Dataset D1 R^2 on trained D1: 0.17245875339912875 .................................................. Pipeline #12: Score on D2: 0.11795572539859911 | D1-D2 diff: 2.35005179133754 Pipeline steps: SelectPercentile(percentile=80), HeterosisEncoder(), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=5, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1386369996159964 Holdout data R^2 trained on entire dataset(80%): 0.12435144896721173 Dataset D1 R^2 on trained D1: 0.15074186365275954 .................................................. Pipeline #13: Score on D2: 0.09042424288609818 | D1-D2 diff: 3.239382016047848 Pipeline steps: SelectPercentile(percentile=30), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09565624966027275 Holdout data R^2 trained on entire dataset(80%): 0.09470207450518797 Dataset D1 R^2 on trained D1: 0.09950561209961939 .................................................. Pipeline #14: Score on D2: 0.08637210235203885 | D1-D2 diff: 3.5069415282337815 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08984567535359544 Holdout data R^2 trained on entire dataset(80%): 0.08619380614950267 Dataset D1 R^2 on trained D1: 0.09298338767604586 .................................................. Pipeline #15: Score on D2: 0.04867344746978708 | D1-D2 diff: 4.30239161582635 Pipeline steps: SelectPercentile(percentile=90), DecisionTreeRegressor(max_depth=2, min_samples_leaf=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05066496039287349 Holdout data R^2 trained on entire dataset(80%): 0.044260451659357436 Dataset D1 R^2 on trained D1: 0.051591951153166504 .................................................. Pipeline #16: Score on D2: 0.04867344746978697 | D1-D2 diff: 4.302391615826391 Pipeline steps: SelectPercentile(percentile=95), DecisionTreeRegressor(max_depth=2, min_samples_leaf=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05066496039287349 Holdout data R^2 trained on entire dataset(80%): 0.044260451659357436 Dataset D1 R^2 on trained D1: 0.05159195115316628 .................................................. Pipeline #17: Score on D2: 0.03657725921554544 | D1-D2 diff: 4.4343073636394 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), SelectPercentile(percentile=85), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.038202290736077504 Holdout data R^2 trained on entire dataset(80%): 0.05439255205489424 Dataset D1 R^2 on trained D1: 0.03916366598547116 .................................................. Pipeline #18: Score on D2: 0.031659849194455525 | D1-D2 diff: 5.0158579995516215 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03252279584647144 Holdout data R^2 trained on entire dataset(80%): 0.040831985722540654 Dataset D1 R^2 on trained D1: 0.0332397108844984 .................................................. Pipeline #19: Score on D2: 0.0236050200545741 | D1-D2 diff: 5.402916377605068 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.023030656997982457 Holdout data R^2 trained on entire dataset(80%): 0.034387046577685654 Dataset D1 R^2 on trained D1: 0.02243150945567851 .................................................. Pipeline #20: Score on D2: 0.009179090116651079 | D1-D2 diff: 11.442166828468554 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), SelectPercentile(percentile=20), DecisionTreeRegressor(max_depth=3, min_samples_leaf=3, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.010314216976863633 Holdout data R^2 trained on entire dataset(80%): 0.02374258257372297 Dataset D1 R^2 on trained D1: 0.009237430180190853 .................................................. Pipeline #21: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=10, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 27 - 8 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.17285 Best score on D1-D2 diff: 12.35154 Gen 2 - Best score on D2: 0.17488 Best score on D1-D2 diff: 13.48015 Gen 3 - Best score on D2: 0.17671 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.17671 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.17671 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.17921 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.18192 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.18192 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.18280 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.18396 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.18396 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.18437 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.18777 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.004797121279856986 Entire dataset(80%) R^2 trained on data (80%): 0.007527791424385311 Holdout R^2 (20%) trained on data (80%): -0.01586822304142066 Dataset D1 R^2 on trained D1: 0.012057583976985864 Combined Dataset (100%) R^2 trained on combined data (100%): 0.00426755244424426 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.18776837784440248 | D1-D2 diff: 1.4096763175686868 Pipeline steps: VarianceThreshold(threshold=0.1), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=2, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.440090909106973 Holdout data R^2 trained on entire dataset(80%): 0.22114935294015403 Dataset D1 R^2 on trained D1: 0.4410025937041311 .................................................. Pipeline #2: Score on D2: 0.18437297603868874 | D1-D2 diff: 1.5978240135261201 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=3, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.339462153206047 Holdout data R^2 trained on entire dataset(80%): 0.2079344907270647 Dataset D1 R^2 on trained D1: 0.33779376955415075 .................................................. Pipeline #3: Score on D2: 0.18396480257873737 | D1-D2 diff: 1.6120826556318186 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=4, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3343385155781239 Holdout data R^2 trained on entire dataset(80%): 0.20985904370230035 Dataset D1 R^2 on trained D1: 0.3320292459355124 .................................................. Pipeline #4: Score on D2: 0.18279897851306282 | D1-D2 diff: 1.616979598902709 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=4, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3342461404943351 Holdout data R^2 trained on entire dataset(80%): 0.21023423769123173 Dataset D1 R^2 on trained D1: 0.3290779297149268 .................................................. Pipeline #5: Score on D2: 0.17637949826526045 | D1-D2 diff: 1.664052641588654 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=8, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.309008384688062 Holdout data R^2 trained on entire dataset(80%): 0.1996938620395441 Dataset D1 R^2 on trained D1: 0.30679576271724107 .................................................. Pipeline #6: Score on D2: 0.1755684742602971 | D1-D2 diff: 1.6942458563776033 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2988872241324577 Holdout data R^2 trained on entire dataset(80%): 0.19473537579277334 Dataset D1 R^2 on trained D1: 0.2969337009361581 .................................................. Pipeline #7: Score on D2: 0.17305871194592093 | D1-D2 diff: 1.7251029279822487 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=13, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2949336757059724 Holdout data R^2 trained on entire dataset(80%): 0.19920060201836998 Dataset D1 R^2 on trained D1: 0.2859706731413133 .................................................. Pipeline #8: Score on D2: 0.17296087888286438 | D1-D2 diff: 1.7464280831375862 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=13, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2885949508104976 Holdout data R^2 trained on entire dataset(80%): 0.19756814948784285 Dataset D1 R^2 on trained D1: 0.2804580844102642 .................................................. Pipeline #9: Score on D2: 0.17257079744562642 | D1-D2 diff: 1.7672583143000271 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=11, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2809051948474095 Holdout data R^2 trained on entire dataset(80%): 0.19242893724528565 Dataset D1 R^2 on trained D1: 0.27508873626466257 .................................................. Pipeline #10: Score on D2: 0.16998891865444876 | D1-D2 diff: 1.8187727701085334 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27129189076681837 Holdout data R^2 trained on entire dataset(80%): 0.19808788123391474 Dataset D1 R^2 on trained D1: 0.26137629851069943 .................................................. Pipeline #11: Score on D2: 0.16956827532744 | D1-D2 diff: 1.8342611464129857 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=17, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26753479584729034 Holdout data R^2 trained on entire dataset(80%): 0.19199398146862445 Dataset D1 R^2 on trained D1: 0.2579078558147513 .................................................. Pipeline #12: Score on D2: 0.1671928688630484 | D1-D2 diff: 1.8346087232797441 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=17, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2689184379163675 Holdout data R^2 trained on entire dataset(80%): 0.19778493369286076 Dataset D1 R^2 on trained D1: 0.25546552266562983 .................................................. Pipeline #13: Score on D2: 0.16714087486752027 | D1-D2 diff: 1.8753370272774565 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25927568552606617 Holdout data R^2 trained on entire dataset(80%): 0.1887147178180646 Dataset D1 R^2 on trained D1: 0.24799137034655183 .................................................. Pipeline #14: Score on D2: 0.1665459908158332 | D1-D2 diff: 1.8923915071509367 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RecessiveEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=15, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2549542222572606 Holdout data R^2 trained on entire dataset(80%): 0.18592427170568737 Dataset D1 R^2 on trained D1: 0.24452110847374997 .................................................. Pipeline #15: Score on D2: 0.16402377442851723 | D1-D2 diff: 1.9074788895234782 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=20, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25551301132957527 Holdout data R^2 trained on entire dataset(80%): 0.19029587433340278 Dataset D1 R^2 on trained D1: 0.23956100167808714 .................................................. Pipeline #16: Score on D2: 0.16271258215519424 | D1-D2 diff: 1.9450007365352928 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24563608025110795 Holdout data R^2 trained on entire dataset(80%): 0.18127894889587826 Dataset D1 R^2 on trained D1: 0.23258743712978114 .................................................. Pipeline #17: Score on D2: 0.16173836204469627 | D1-D2 diff: 1.996357159914159 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=19, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2362176497489017 Holdout data R^2 trained on entire dataset(80%): 0.18152435419043744 Dataset D1 R^2 on trained D1: 0.2246957981147938 .................................................. Pipeline #18: Score on D2: 0.15341548708879238 | D1-D2 diff: 2.0032050863215742 Pipeline steps: VarianceThreshold(threshold=0.2), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=19, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2261102023119883 Holdout data R^2 trained on entire dataset(80%): 0.17444320566203242 Dataset D1 R^2 on trained D1: 0.21551645125889674 .................................................. Pipeline #19: Score on D2: 0.13777650589672585 | D1-D2 diff: 2.0080356796239234 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=9, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21535587784478472 Holdout data R^2 trained on entire dataset(80%): 0.1548520979781255 Dataset D1 R^2 on trained D1: 0.19928205483295203 .................................................. Pipeline #20: Score on D2: 0.13288166815453117 | D1-D2 diff: 2.129372023254342 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=95), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RecessiveEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19414745458130722 Holdout data R^2 trained on entire dataset(80%): 0.15066944717146324 Dataset D1 R^2 on trained D1: 0.18152169711329902 .................................................. Pipeline #21: Score on D2: 0.12364189121310365 | D1-D2 diff: 2.1796158016250575 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=13, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18615997111440774 Holdout data R^2 trained on entire dataset(80%): 0.1475608869660784 Dataset D1 R^2 on trained D1: 0.16794969348452926 .................................................. Pipeline #22: Score on D2: 0.11975572264913148 | D1-D2 diff: 2.392933296792096 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=70), DecisionTreeRegressor(max_depth=4, min_samples_leaf=6, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12434978381156514 Holdout data R^2 trained on entire dataset(80%): 0.10468101458941792 Dataset D1 R^2 on trained D1: 0.15025416287116777 .................................................. Pipeline #23: Score on D2: 0.10823723891925963 | D1-D2 diff: 2.426009980995835 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15992622939165924 Holdout data R^2 trained on entire dataset(80%): 0.12950137470369294 Dataset D1 R^2 on trained D1: 0.13710610107514187 .................................................. Pipeline #24: Score on D2: 0.10057755132505108 | D1-D2 diff: 3.0251068131061443 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11074746236403521 Holdout data R^2 trained on entire dataset(80%): 0.10933378870224797 Dataset D1 R^2 on trained D1: 0.11251845363042534 .................................................. Pipeline #25: Score on D2: 0.08523767734125998 | D1-D2 diff: 3.321244474644455 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08952662851035442 Holdout data R^2 trained on entire dataset(80%): 0.08622323521977404 Dataset D1 R^2 on trained D1: 0.09345625418874526 .................................................. Pipeline #26: Score on D2: 0.051969707751231 | D1-D2 diff: 3.782624488674825 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), FeatureEncodingFrequencySelector(threshold=0.35), RecessiveEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05565572501074745 Holdout data R^2 trained on entire dataset(80%): 0.058016622176932264 Dataset D1 R^2 on trained D1: 0.056854286105768925 .................................................. Pipeline #27: Score on D2: 0.04515338171121863 | D1-D2 diff: 4.403573685383726 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.044986910780409284 Holdout data R^2 trained on entire dataset(80%): 0.05503070364493201 Dataset D1 R^2 on trained D1: 0.042494010710279806 .................................................. Pipeline #28: Score on D2: 0.044914098311441664 | D1-D2 diff: 4.583739170961737 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.045176373031454276 Holdout data R^2 trained on entire dataset(80%): 0.05568401193096484 Dataset D1 R^2 on trained D1: 0.042648826023089614 .................................................. Pipeline #29: Score on D2: 0.03986865076343249 | D1-D2 diff: 6.0816822844532785 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.039667989164889006 Holdout data R^2 trained on entire dataset(80%): 0.052984661598875715 Dataset D1 R^2 on trained D1: 0.03913767144948754 .................................................. Pipeline #30: Score on D2: 0.03975979569208732 | D1-D2 diff: 7.4960793401261245 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04082599526105557 Holdout data R^2 trained on entire dataset(80%): 0.052472940497541476 Dataset D1 R^2 on trained D1: 0.04007650680452057 .................................................. Pipeline #31: Score on D2: 0.0313352354945754 | D1-D2 diff: 11.22238495921987 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=13, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03131289521143099 Holdout data R^2 trained on entire dataset(80%): 0.029959164912579217 Dataset D1 R^2 on trained D1: 0.03127218923126873 .................................................. Pipeline #32: Score on D2: 0.031320014308591326 | D1-D2 diff: 12.351543538013058 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.031316532891997984 Holdout data R^2 trained on entire dataset(80%): 0.03002870948599823 Dataset D1 R^2 on trained D1: 0.031277049283894964 .................................................. Pipeline #33: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: RecessiveEncoder(), OverDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=12, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 27 - 9 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.17904 Best score on D1-D2 diff: 4.41803 Gen 2 - Best score on D2: 0.18352 Best score on D1-D2 diff: 4.41803 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.01021582845644331 Entire dataset(80%) R^2 trained on data (80%): 0.007117003152831369 Holdout R^2 (20%) trained on data (80%): -0.011673002494141382 Dataset D1 R^2 on trained D1: 0.013902952893817155 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004637977029731344 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.18352398190279828 | D1-D2 diff: 1.398998611659457 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.1, min_samples_leaf=3, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.44412990019808185 Holdout data R^2 trained on entire dataset(80%): 0.19870439615802005 Dataset D1 R^2 on trained D1: 0.4445782908079421 .................................................. Pipeline #2: Score on D2: 0.17904109046834227 | D1-D2 diff: 1.543287877982021 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=3, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3555544432846317 Holdout data R^2 trained on entire dataset(80%): 0.19370913885403307 Dataset D1 R^2 on trained D1: 0.3553248159861211 .................................................. Pipeline #3: Score on D2: 0.1777965464679524 | D1-D2 diff: 1.7303304425755335 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2870848839394481 Holdout data R^2 trained on entire dataset(80%): 0.19379532331216587 Dataset D1 R^2 on trained D1: 0.28935020179229864 .................................................. Pipeline #4: Score on D2: 0.14099628300884537 | D1-D2 diff: 2.0741323441111765 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.1, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21265494216460712 Holdout data R^2 trained on entire dataset(80%): 0.1657539168418931 Dataset D1 R^2 on trained D1: 0.19502866880224268 .................................................. Pipeline #5: Score on D2: 0.11230244979226722 | D1-D2 diff: 2.187146721529786 Pipeline steps: UnderDominanceEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1695671586448878 Holdout data R^2 trained on entire dataset(80%): 0.13186196967975083 Dataset D1 R^2 on trained D1: 0.15600314318428388 .................................................. Pipeline #6: Score on D2: 0.08938002582919646 | D1-D2 diff: 2.4040306325930536 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10561068683724018 Holdout data R^2 trained on entire dataset(80%): 0.1020005266536036 Dataset D1 R^2 on trained D1: 0.11931921339588669 .................................................. Pipeline #7: Score on D2: 0.0779396002864412 | D1-D2 diff: 2.7126836005675052 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=50), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08926303587886575 Holdout data R^2 trained on entire dataset(80%): 0.10110857606188517 Dataset D1 R^2 on trained D1: 0.09640690148260267 .................................................. Pipeline #8: Score on D2: 0.07256921470844202 | D1-D2 diff: 2.9166528584322555 Pipeline steps: VarianceThreshold(threshold=0.25), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08158285067597437 Holdout data R^2 trained on entire dataset(80%): 0.07446113690131795 Dataset D1 R^2 on trained D1: 0.08638771878669527 .................................................. Pipeline #9: Score on D2: 0.06052990231277955 | D1-D2 diff: 2.9191488903525857 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06818109577892073 Holdout data R^2 trained on entire dataset(80%): 0.082160774850769 Dataset D1 R^2 on trained D1: 0.07430120466793899 .................................................. Pipeline #10: Score on D2: 0.05556777413538294 | D1-D2 diff: 3.475204263886358 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.060748138132491136 Holdout data R^2 trained on entire dataset(80%): 0.0683916584156321 Dataset D1 R^2 on trained D1: 0.06242389798761727 .................................................. Pipeline #11: Score on D2: 0.0507240641304364 | D1-D2 diff: 4.123236166803186 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=15, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05251512754098109 Holdout data R^2 trained on entire dataset(80%): 0.06234963906486779 Dataset D1 R^2 on trained D1: 0.05418383356459344 .................................................. Pipeline #12: Score on D2: 0.009916686101958061 | D1-D2 diff: 4.418034018349015 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.027042393349901994 Holdout data R^2 trained on entire dataset(80%): 0.04363824544640271 Dataset D1 R^2 on trained D1: 0.012541410920026141 .................................................. ************************************************************************************** Random Seed 28 - 0 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.00046 Best score on D1-D2 diff: 5.07154 Gen 2 - Best score on D2: 0.00046 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00046 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00117 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00117 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00117 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.00147 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.011665049760223978 Entire dataset(80%) R^2 trained on data (80%): 0.004347658565957091 Holdout R^2 (20%) trained on data (80%): -0.005226705978988111 Dataset D1 R^2 on trained D1: 0.006202056339530304 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003372855625611293 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.0014715464138508327 | D1-D2 diff: 6.332700509107936 Pipeline steps: RecessiveEncoder(), SelectPercentile(percentile=70), VarianceThreshold(threshold=0.1), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.003711716780436136 Holdout data R^2 trained on entire dataset(80%): -0.004765405707261028 Dataset D1 R^2 on trained D1: 0.0020933370849146593 .................................................. Pipeline #2: Score on D2: 0.000922029987623807 | D1-D2 diff: 10.261510501130118 Pipeline steps: SelectPercentile(percentile=50), FeatureEncodingFrequencySelector(threshold=0.2), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0018063805433958802 Holdout data R^2 trained on entire dataset(80%): -0.0037109651680389266 Dataset D1 R^2 on trained D1: 0.0010122192477830527 .................................................. Pipeline #3: Score on D2: -3.2015594250811574e-05 | D1-D2 diff: 13.480151627526027 Pipeline steps: DominantEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021467281316e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516763899763 Dataset D1 R^2 on trained D1: -1.73106056000627e-06 .................................................. ************************************************************************************** Random Seed 28 - 1 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.08929 Best score on D1-D2 diff: 6.74555 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.002453636810862303 Entire dataset(80%) R^2 trained on data (80%): 0.004983346716332715 Holdout R^2 (20%) trained on data (80%): -0.006726337957656092 Dataset D1 R^2 on trained D1: 0.004773180794318099 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003579962384806956 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.08929223797213914 | D1-D2 diff: 2.7633452754397547 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10081247049175779 Holdout data R^2 trained on entire dataset(80%): 0.06291696074858455 Dataset D1 R^2 on trained D1: 0.10644205098615966 .................................................. Pipeline #2: Score on D2: 0.041904189028707295 | D1-D2 diff: 4.106267453627876 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04375080314176549 Holdout data R^2 trained on entire dataset(80%): 0.03367772059260388 Dataset D1 R^2 on trained D1: 0.045421502444339845 .................................................. Pipeline #3: Score on D2: 0.0018579136700666021 | D1-D2 diff: 5.093940714963183 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00478882109357659 Holdout data R^2 trained on entire dataset(80%): -0.004089100909970833 Dataset D1 R^2 on trained D1: 0.003343112010414928 .................................................. Pipeline #4: Score on D2: 0.0009407289529683727 | D1-D2 diff: 5.923937672862845 Pipeline steps: SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0016221025509268738 Holdout data R^2 trained on entire dataset(80%): -0.002166621999391216 Dataset D1 R^2 on trained D1: 0.0017527327830976214 .................................................. Pipeline #5: Score on D2: 0.0006445448226812811 | D1-D2 diff: 6.745550602125897 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0009273638565640008 Holdout data R^2 trained on entire dataset(80%): -0.0014781078226024924 Dataset D1 R^2 on trained D1: 0.0011275261958498817 .................................................. ************************************************************************************** Random Seed 28 - 2 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.09311 Best score on D1-D2 diff: 5.66722 Gen 2 - Best score on D2: 0.09311 Best score on D1-D2 diff: 11.96036 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0031017940293629476 Entire dataset(80%) R^2 trained on data (80%): 0.004123303437041814 Holdout R^2 (20%) trained on data (80%): -0.002059007210116981 Dataset D1 R^2 on trained D1: 0.0041703996200763704 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0036633394005984865 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.09310567321093322 | D1-D2 diff: 3.322855799803028 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0045159639966900755 Holdout data R^2 trained on entire dataset(80%): -0.004333230167267388 Dataset D1 R^2 on trained D1: 0.10130832018246849 .................................................. Pipeline #2: Score on D2: 0.0008547579309582387 | D1-D2 diff: 5.311725147868958 Pipeline steps: SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0020519233047634478 Holdout data R^2 trained on entire dataset(80%): -0.0011673518131727345 Dataset D1 R^2 on trained D1: 0.0021109545537547625 .................................................. Pipeline #3: Score on D2: 0.0006847457852706684 | D1-D2 diff: 5.667223470817012 Pipeline steps: SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0014409299225834893 Holdout data R^2 trained on entire dataset(80%): -0.0021433051368904277 Dataset D1 R^2 on trained D1: 0.0016541806785650426 .................................................. Pipeline #4: Score on D2: -4.886780840163141e-05 | D1-D2 diff: 11.960360895673112 Pipeline steps: RecessiveEncoder(), DominantEncoder(), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0 Holdout data R^2 trained on entire dataset(80%): -0.0012028184006112053 Dataset D1 R^2 on trained D1: 0.0 .................................................. ************************************************************************************** Random Seed 28 - 3 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.09311 Best score on D1-D2 diff: 5.66722 Gen 2 - Best score on D2: 0.09311 Best score on D1-D2 diff: 11.96036 Gen 1 - Best score on D2: 0.11306 Best score on D1-D2 diff: 4.35454 Gen 2 - Best score on D2: 0.11317 Best score on D1-D2 diff: 4.47701 Gen 3 - Best score on D2: 0.12044 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.12044 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.12044 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.12044 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.12044 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.12044 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.00012233166911546078 Entire dataset(80%) R^2 trained on data (80%): 0.006100057114506674 Holdout R^2 (20%) trained on data (80%): -0.006320193156879483 Dataset D1 R^2 on trained D1: 0.005973293376398181 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004282267352262337 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.12044366534225337 | D1-D2 diff: 1.644140677694209 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=13, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.256443049421908 Holdout data R^2 trained on entire dataset(80%): 0.12531826281501202 Dataset D1 R^2 on trained D1: 0.2572934450288218 .................................................. Pipeline #2: Score on D2: 0.1197944368099152 | D1-D2 diff: 1.6747304272412662 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=13, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24668152812359723 Holdout data R^2 trained on entire dataset(80%): 0.1270551284011009 Dataset D1 R^2 on trained D1: 0.2469163313309649 .................................................. Pipeline #3: Score on D2: 0.1175223583734416 | D1-D2 diff: 1.8418747748316062 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=20, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20467596002492083 Holdout data R^2 trained on entire dataset(80%): 0.12693877762807326 Dataset D1 R^2 on trained D1: 0.20441031814112265 .................................................. Pipeline #4: Score on D2: 0.11513445345268825 | D1-D2 diff: 1.8476826031835365 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=55), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=13, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20773539833640853 Holdout data R^2 trained on entire dataset(80%): 0.13331159268568293 Dataset D1 R^2 on trained D1: 0.20093509221121997 .................................................. Pipeline #5: Score on D2: 0.10882276083201925 | D1-D2 diff: 2.534106847024732 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09934942612867348 Holdout data R^2 trained on entire dataset(80%): 0.08333947554149934 Dataset D1 R^2 on trained D1: 0.1330721224862016 .................................................. Pipeline #6: Score on D2: 0.09254817502182489 | D1-D2 diff: 3.0400825786475587 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=25), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0035552912383916002 Holdout data R^2 trained on entire dataset(80%): -0.0002526191443599224 Dataset D1 R^2 on trained D1: 0.1042555216615405 .................................................. Pipeline #7: Score on D2: 0.09230105689447576 | D1-D2 diff: 3.1616105616886374 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.003656836832587107 Holdout data R^2 trained on entire dataset(80%): -0.007463576762989144 Dataset D1 R^2 on trained D1: 0.10230949954903634 .................................................. Pipeline #8: Score on D2: 0.025676952038965806 | D1-D2 diff: 3.8563964988346386 Pipeline steps: DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07277139713517522 Holdout data R^2 trained on entire dataset(80%): 0.07954302362589538 Dataset D1 R^2 on trained D1: 0.030198355687237055 .................................................. Pipeline #9: Score on D2: 0.002552252552421197 | D1-D2 diff: 7.153441665018197 Pipeline steps: DominantEncoder(), VarianceThreshold(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.3), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002690751573853123 Holdout data R^2 trained on entire dataset(80%): -0.003269485135914074 Dataset D1 R^2 on trained D1: 0.002170361185877212 .................................................. Pipeline #10: Score on D2: 0.0012862668109622222 | D1-D2 diff: 7.194014655737301 Pipeline steps: SelectPercentile(percentile=40), OverDominanceEncoder(), VarianceThreshold(threshold=0.35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0020033031317194805 Holdout data R^2 trained on entire dataset(80%): -0.0008218392968961652 Dataset D1 R^2 on trained D1: 0.0016596155828406678 .................................................. Pipeline #11: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: DominantEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=6, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 28 - 4 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.14442 Best score on D1-D2 diff: 4.33658 Gen 2 - Best score on D2: 0.14927 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.15491 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15984 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.15984 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.15984 Best score on D1-D2 diff: 11.96036 Gen 7 - Best score on D2: 0.15984 Best score on D1-D2 diff: 11.96036 Gen 8 - Best score on D2: 0.15984 Best score on D1-D2 diff: 11.96036 Gen 9 - Best score on D2: 0.16015 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.16106 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.16106 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.16106 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.16106 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.16139 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.16468 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.16468 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.16468 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.16468 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.16468 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.16468 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.16468 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.16468 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.16494 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.16494 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.16494 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.005594788565307818 Entire dataset(80%) R^2 trained on data (80%): 0.006105630309219956 Holdout R^2 (20%) trained on data (80%): -0.0048946954849971025 Dataset D1 R^2 on trained D1: 0.007918011263747138 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004883931022773291 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.16493600487567905 | D1-D2 diff: 1.7059932904806805 Pipeline steps: SelectPercentile(percentile=80), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.5, min_samples_leaf=16, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2391634840101322 Holdout data R^2 trained on entire dataset(80%): 0.08763909019428251 Dataset D1 R^2 on trained D1: 0.2829927281072524 .................................................. Pipeline #2: Score on D2: 0.16071175299767182 | D1-D2 diff: 1.758341388271996 Pipeline steps: SelectPercentile(percentile=70), VarianceThreshold(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(max_features=0.8500000000000001, min_samples_leaf=14, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16547499113137942 Holdout data R^2 trained on entire dataset(80%): 0.03723973292748051 Dataset D1 R^2 on trained D1: 0.2653251259777668 .................................................. Pipeline #3: Score on D2: 0.16037443191899925 | D1-D2 diff: 1.7673319004512273 Pipeline steps: SelectPercentile(percentile=80), FeatureEncodingFrequencySelector(threshold=0.05), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.5, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21716594333707206 Holdout data R^2 trained on entire dataset(80%): 0.0903744848268363 Dataset D1 R^2 on trained D1: 0.2628752977040426 .................................................. Pipeline #4: Score on D2: 0.15871293837090428 | D1-D2 diff: 1.904494283776947 Pipeline steps: SelectPercentile(percentile=80), FeatureEncodingFrequencySelector(threshold=0.05), HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1988982408114539 Holdout data R^2 trained on entire dataset(80%): 0.09082580986549771 Dataset D1 R^2 on trained D1: 0.2347247889629398 .................................................. Pipeline #5: Score on D2: 0.15541596322297047 | D1-D2 diff: 1.9231574420390547 Pipeline steps: SelectPercentile(percentile=80), FeatureEncodingFrequencySelector(threshold=0.05), HeterosisEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19054596166021076 Holdout data R^2 trained on entire dataset(80%): 0.09458210853351723 Dataset D1 R^2 on trained D1: 0.22851987905102977 .................................................. Pipeline #6: Score on D2: 0.1319172251281484 | D1-D2 diff: 2.4793133806717456 Pipeline steps: SelectPercentile(percentile=70), VarianceThreshold(threshold=0.35), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10605033600200808 Holdout data R^2 trained on entire dataset(80%): 0.054809787382183006 Dataset D1 R^2 on trained D1: 0.15838237166071245 .................................................. Pipeline #7: Score on D2: 0.12947058607122164 | D1-D2 diff: 2.569291449854794 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=75), DecisionTreeRegressor(max_depth=4, min_samples_leaf=3, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14011142356801987 Holdout data R^2 trained on entire dataset(80%): 0.10017421385979841 Dataset D1 R^2 on trained D1: 0.15241867443142287 .................................................. Pipeline #8: Score on D2: 0.1114757369163425 | D1-D2 diff: 3.3213177981538067 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), RecessiveEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09716907623859927 Holdout data R^2 trained on entire dataset(80%): 0.06896300200507588 Dataset D1 R^2 on trained D1: 0.11969358803388841 .................................................. Pipeline #9: Score on D2: 0.044225946829620666 | D1-D2 diff: 4.0569202243251 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04668013854914577 Holdout data R^2 trained on entire dataset(80%): 0.02221845353107854 Dataset D1 R^2 on trained D1: 0.04791754250783875 .................................................. Pipeline #10: Score on D2: 0.023812765415962 | D1-D2 diff: 5.086412408347031 Pipeline steps: UnderDominanceEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), SelectPercentile(percentile=90), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05251788619657605 Holdout data R^2 trained on entire dataset(80%): 0.05617627512892431 Dataset D1 R^2 on trained D1: 0.022318754674769403 .................................................. Pipeline #11: Score on D2: 0.02381276541596178 | D1-D2 diff: 5.086412408347219 Pipeline steps: UnderDominanceEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05251788619657605 Holdout data R^2 trained on entire dataset(80%): 0.05617627512892431 Dataset D1 R^2 on trained D1: 0.022318754674769403 .................................................. Pipeline #12: Score on D2: 0.014891327019974088 | D1-D2 diff: 6.330684548539763 Pipeline steps: DominantEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), RecessiveEncoder(), RecessiveEncoder(), SelectPercentile(percentile=60), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004004254320205347 Holdout data R^2 trained on entire dataset(80%): -0.0016944448963380765 Dataset D1 R^2 on trained D1: 0.01551391008835079 .................................................. Pipeline #13: Score on D2: 0.002141748306856317 | D1-D2 diff: 8.757353537046404 Pipeline steps: SelectPercentile(percentile=15), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002234991454386215 Holdout data R^2 trained on entire dataset(80%): -0.0001855654395990225 Dataset D1 R^2 on trained D1: 0.0019717249958326466 .................................................. Pipeline #14: Score on D2: 0.00204557706331987 | D1-D2 diff: 10.930944175161564 Pipeline steps: SelectPercentile(percentile=15), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002096950865866476 Holdout data R^2 trained on entire dataset(80%): 0.0007607994233234106 Dataset D1 R^2 on trained D1: 0.0019755333285871313 .................................................. Pipeline #15: Score on D2: 0.002045577063319648 | D1-D2 diff: 10.930944175170227 Pipeline steps: SelectPercentile(percentile=15), HeterosisEncoder(), SelectPercentile(percentile=75), DecisionTreeRegressor(max_depth=4, min_samples_leaf=6, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002096950865866476 Holdout data R^2 trained on entire dataset(80%): 0.0007607994233234106 Dataset D1 R^2 on trained D1: 0.0019755333285871313 .................................................. Pipeline #16: Score on D2: 0.0018886080179379983 | D1-D2 diff: 11.236932052701244 Pipeline steps: SelectPercentile(percentile=15), HeterosisEncoder(), SelectPercentile(percentile=70), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0019620005883579372 Holdout data R^2 trained on entire dataset(80%): 0.0005144337886583417 Dataset D1 R^2 on trained D1: 0.0019513284412543408 .................................................. Pipeline #17: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: RecessiveEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RandomForestRegressor(max_features=0.8, min_samples_leaf=2, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 28 - 5 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.16258 Best score on D1-D2 diff: 5.22425 Gen 2 - Best score on D2: 0.16258 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.16505 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.16505 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.17043 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.17043 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.17319 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.17451 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.17451 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.002326832222649733 Entire dataset(80%) R^2 trained on data (80%): 0.004447575727700048 Holdout R^2 (20%) trained on data (80%): -0.004809455395247131 Dataset D1 R^2 on trained D1: 0.0039540458182030225 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0034608355567702365 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.17450630837063186 | D1-D2 diff: 1.4335321526689768 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.41457923214623893 Holdout data R^2 trained on entire dataset(80%): 0.16312910706525396 Dataset D1 R^2 on trained D1: 0.4113000649053018 .................................................. Pipeline #2: Score on D2: 0.17033757860771692 | D1-D2 diff: 1.4834343500859004 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=8, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3771264110591185 Holdout data R^2 trained on entire dataset(80%): 0.14895930850580208 Dataset D1 R^2 on trained D1: 0.3768407259167841 .................................................. Pipeline #3: Score on D2: 0.16832926613371524 | D1-D2 diff: 1.4837796259165172 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=8, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.37570581686793825 Holdout data R^2 trained on entire dataset(80%): 0.15068870332247097 Dataset D1 R^2 on trained D1: 0.3746402672214705 .................................................. Pipeline #4: Score on D2: 0.16812339146748567 | D1-D2 diff: 1.7592422381735524 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.45, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2839400190655892 Holdout data R^2 trained on entire dataset(80%): 0.1528922768755112 Dataset D1 R^2 on trained D1: 0.27252265275131415 .................................................. Pipeline #5: Score on D2: 0.16306623851148938 | D1-D2 diff: 1.7823127532859757 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=20, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27146951822667953 Holdout data R^2 trained on entire dataset(80%): 0.15227199529318092 Dataset D1 R^2 on trained D1: 0.2621641133815419 .................................................. Pipeline #6: Score on D2: 0.16190804312898854 | D1-D2 diff: 1.785026314956905 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=20, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26999463500016174 Holdout data R^2 trained on entire dataset(80%): 0.14969858579282358 Dataset D1 R^2 on trained D1: 0.26040470415587513 .................................................. Pipeline #7: Score on D2: 0.16081043720132926 | D1-D2 diff: 1.8844148801870022 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=19, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2546936476553713 Holdout data R^2 trained on entire dataset(80%): 0.153999979954198 Dataset D1 R^2 on trained D1: 0.24011421934986876 .................................................. Pipeline #8: Score on D2: 0.1542730768930205 | D1-D2 diff: 1.9044393821831438 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(max_features=0.45, min_samples_leaf=19, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24082178131931142 Holdout data R^2 trained on entire dataset(80%): 0.14856577328599507 Dataset D1 R^2 on trained D1: 0.23029369300890645 .................................................. Pipeline #9: Score on D2: 0.15366549998642265 | D1-D2 diff: 1.92846430147408 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), OverDominanceEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=19, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24099322818096358 Holdout data R^2 trained on entire dataset(80%): 0.15151261904398072 Dataset D1 R^2 on trained D1: 0.22596804498887402 .................................................. Pipeline #10: Score on D2: 0.15277357705309558 | D1-D2 diff: 1.9452252784208255 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23924422256116284 Holdout data R^2 trained on entire dataset(80%): 0.15378406696477542 Dataset D1 R^2 on trained D1: 0.2226161743444215 .................................................. Pipeline #11: Score on D2: 0.15077575719918213 | D1-D2 diff: 1.9534228128314768 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), VarianceThreshold(threshold=0.05), RandomForestRegressor(max_features=0.4, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23538445995589297 Holdout data R^2 trained on entire dataset(80%): 0.15023911617372598 Dataset D1 R^2 on trained D1: 0.21945333647782617 .................................................. Pipeline #12: Score on D2: 0.14515520362882472 | D1-D2 diff: 1.9849456783560397 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), VarianceThreshold(threshold=0.05), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2192399543291218 Holdout data R^2 trained on entire dataset(80%): 0.14280740518400314 Dataset D1 R^2 on trained D1: 0.20957294537958482 .................................................. Pipeline #13: Score on D2: 0.14370377604666273 | D1-D2 diff: 2.130008114566979 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=35), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.053213639538904234 Holdout data R^2 trained on entire dataset(80%): 0.007740748619393778 Dataset D1 R^2 on trained D1: 0.19228572890127293 .................................................. Pipeline #14: Score on D2: 0.12581326112203328 | D1-D2 diff: 2.441656369518257 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), SelectPercentile(percentile=75), RecessiveEncoder(), DecisionTreeRegressor(max_depth=5, min_samples_leaf=19, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14992661664041562 Holdout data R^2 trained on entire dataset(80%): 0.10520313743466125 Dataset D1 R^2 on trained D1: 0.15394922701994262 .................................................. Pipeline #15: Score on D2: 0.10277125617011795 | D1-D2 diff: 2.6282486306632546 Pipeline steps: SelectPercentile(percentile=40), HeterosisEncoder(), SelectPercentile(percentile=35), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=15, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0028663151317148694 Holdout data R^2 trained on entire dataset(80%): -0.0004880075913220594 Dataset D1 R^2 on trained D1: 0.12372850224190446 .................................................. Pipeline #16: Score on D2: 0.09611804328619422 | D1-D2 diff: 5.224248225484633 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), SelectPercentile(percentile=75), RecessiveEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=15, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09652002597645237 Holdout data R^2 trained on entire dataset(80%): 0.10291599894665326 Dataset D1 R^2 on trained D1: 0.09477557268293735 .................................................. Pipeline #17: Score on D2: 0.09611804328619411 | D1-D2 diff: 5.224248225484741 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09652002597645237 Holdout data R^2 trained on entire dataset(80%): 0.10291599894665326 Dataset D1 R^2 on trained D1: 0.09477557268293735 .................................................. Pipeline #18: Score on D2: 0.096118043286194 | D1-D2 diff: 5.224248225484849 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=90), DecisionTreeRegressor(max_depth=2, min_samples_leaf=4, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09652002597645237 Holdout data R^2 trained on entire dataset(80%): 0.10291599894665326 Dataset D1 R^2 on trained D1: 0.09477557268293735 .................................................. Pipeline #19: Score on D2: 0.015563377384780419 | D1-D2 diff: 5.439277373137713 Pipeline steps: DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.016306472748087386 Holdout data R^2 trained on entire dataset(80%): 0.010655277662708573 Dataset D1 R^2 on trained D1: 0.014420932714373502 .................................................. Pipeline #20: Score on D2: 0.015155474825576998 | D1-D2 diff: 5.604127476283979 Pipeline steps: DominantEncoder(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), SelectPercentile(percentile=75), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.016300940063647573 Holdout data R^2 trained on entire dataset(80%): 0.01075378411160488 Dataset D1 R^2 on trained D1: 0.014141638196761508 .................................................. Pipeline #21: Score on D2: 0.014809679067495085 | D1-D2 diff: 6.0402704021865885 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=55), FeatureEncodingFrequencySelector(threshold=0.2), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.016214914500978894 Holdout data R^2 trained on entire dataset(80%): 0.010783677923234758 Dataset D1 R^2 on trained D1: 0.014058446380572631 .................................................. Pipeline #22: Score on D2: 0.0022995544608682694 | D1-D2 diff: 6.459049428651083 Pipeline steps: DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00305673609048418 Holdout data R^2 trained on entire dataset(80%): 7.232492773245891e-05 Dataset D1 R^2 on trained D1: 0.0017250074522768832 .................................................. Pipeline #23: Score on D2: 0.0018928179485953045 | D1-D2 diff: 8.33092952167991 Pipeline steps: DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), VarianceThreshold(threshold=0.2), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0021466445126732125 Holdout data R^2 trained on entire dataset(80%): 0.0006696693466269332 Dataset D1 R^2 on trained D1: 0.0016852185178517498 .................................................. Pipeline #24: Score on D2: 0.0006649096933541987 | D1-D2 diff: 9.248727981179888 Pipeline steps: SelectPercentile(percentile=10), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=18, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0012339385351669563 Holdout data R^2 trained on entire dataset(80%): -0.0007115600025258129 Dataset D1 R^2 on trained D1: 0.0008015793094360291 .................................................. Pipeline #25: Score on D2: 0.0006618895012675541 | D1-D2 diff: 9.51048641568591 Pipeline steps: SelectPercentile(percentile=10), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0012262134765114174 Holdout data R^2 trained on entire dataset(80%): -0.0006961068903259537 Dataset D1 R^2 on trained D1: 0.0007841226731689632 .................................................. Pipeline #26: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=5, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 28 - 6 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.18247 Best score on D1-D2 diff: 5.22425 Gen 2 - Best score on D2: 0.18336 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.18460 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.18460 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.18750 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.18750 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.18750 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.18750 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.18750 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.19233 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.19233 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.19233 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.19233 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.19233 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.19497 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.004071697544616226 Entire dataset(80%) R^2 trained on data (80%): 0.004134928034137175 Holdout R^2 (20%) trained on data (80%): -0.004901461189225742 Dataset D1 R^2 on trained D1: 0.004621373792324079 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0032139794206910155 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.19497171903410937 | D1-D2 diff: 1.4844609652553589 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.7000000000000001, min_samples_leaf=3, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.41637674393443824 Holdout data R^2 trained on entire dataset(80%): 0.18976762636817834 Dataset D1 R^2 on trained D1: 0.4009042095188793 .................................................. Pipeline #2: Score on D2: 0.1941982408491798 | D1-D2 diff: 1.563157632072619 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3774018866520521 Holdout data R^2 trained on entire dataset(80%): 0.19269281802231064 Dataset D1 R^2 on trained D1: 0.36168824697267576 .................................................. Pipeline #3: Score on D2: 0.1912500896787671 | D1-D2 diff: 1.5808878813625156 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.37284706479903407 Holdout data R^2 trained on entire dataset(80%): 0.19721256209577842 Dataset D1 R^2 on trained D1: 0.3513517068963903 .................................................. Pipeline #4: Score on D2: 0.18691228399137239 | D1-D2 diff: 1.7053646534870148 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.55, min_samples_leaf=17, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.32815218691411263 Holdout data R^2 trained on entire dataset(80%): 0.1888371323445146 Dataset D1 R^2 on trained D1: 0.30514317729142404 .................................................. Pipeline #5: Score on D2: 0.18559209283957367 | D1-D2 diff: 1.7157279783225023 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.55, min_samples_leaf=17, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3257456637435002 Holdout data R^2 trained on entire dataset(80%): 0.18717955069099124 Dataset D1 R^2 on trained D1: 0.30099221447959046 .................................................. Pipeline #6: Score on D2: 0.1846117837522664 | D1-D2 diff: 1.7175645538464173 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.5, min_samples_leaf=17, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.32685073974687984 Holdout data R^2 trained on entire dataset(80%): 0.18728112348321568 Dataset D1 R^2 on trained D1: 0.299519111474684 .................................................. Pipeline #7: Score on D2: 0.18443431065563953 | D1-D2 diff: 1.720463789270801 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.6000000000000001, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3229251909223749 Holdout data R^2 trained on entire dataset(80%): 0.18720878555057563 Dataset D1 R^2 on trained D1: 0.29856905077912255 .................................................. Pipeline #8: Score on D2: 0.18335607790922692 | D1-D2 diff: 1.73906842736295 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.4, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3116549354331768 Holdout data R^2 trained on entire dataset(80%): 0.18896861138875576 Dataset D1 R^2 on trained D1: 0.2926845598075918 .................................................. Pipeline #9: Score on D2: 0.1821580948981435 | D1-D2 diff: 1.7737270449649905 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.6000000000000001, min_samples_leaf=20, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3092830828206363 Holdout data R^2 trained on entire dataset(80%): 0.18726717652690572 Dataset D1 R^2 on trained D1: 0.2831886753192189 .................................................. Pipeline #10: Score on D2: 0.18104089090956454 | D1-D2 diff: 1.8190853389977815 Pipeline steps: VarianceThreshold(threshold=0.15), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=1.0, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30113292611451625 Holdout data R^2 trained on entire dataset(80%): 0.17411367114852783 Dataset D1 R^2 on trained D1: 0.2723654754923811 .................................................. Pipeline #11: Score on D2: 0.17839181778305757 | D1-D2 diff: 1.822860335697779 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.9000000000000001, min_samples_leaf=16, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2979144863584914 Holdout data R^2 trained on entire dataset(80%): 0.17671057457983075 Dataset D1 R^2 on trained D1: 0.26896224572031957 .................................................. Pipeline #12: Score on D2: 0.17832637705622767 | D1-D2 diff: 1.8288325089504562 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=15, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2881779229726146 Holdout data R^2 trained on entire dataset(80%): 0.17975937333358694 Dataset D1 R^2 on trained D1: 0.2677195325645876 .................................................. Pipeline #13: Score on D2: 0.17828239842076832 | D1-D2 diff: 1.866651619247448 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.9000000000000001, min_samples_leaf=17, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.292772531176527 Holdout data R^2 trained on entire dataset(80%): 0.17814716193324542 Dataset D1 R^2 on trained D1: 0.2606481971914283 .................................................. Pipeline #14: Score on D2: 0.17748667596508338 | D1-D2 diff: 1.9526970411758349 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), SelectPercentile(percentile=80), SelectPercentile(percentile=95), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23716298120853496 Holdout data R^2 trained on entire dataset(80%): 0.14993008095692917 Dataset D1 R^2 on trained D1: 0.24626641555872664 .................................................. Pipeline #15: Score on D2: 0.17598726612525073 | D1-D2 diff: 2.0026892864534496 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=80), RandomForestRegressor(max_features=0.55, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23805138058299835 Holdout data R^2 trained on entire dataset(80%): 0.14966195601281895 Dataset D1 R^2 on trained D1: 0.2381522323275812 .................................................. Pipeline #16: Score on D2: 0.16823004238719153 | D1-D2 diff: 2.0102818732807237 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2530946352532293 Holdout data R^2 trained on entire dataset(80%): 0.17514690940968625 Dataset D1 R^2 on trained D1: 0.22946115817232626 .................................................. Pipeline #17: Score on D2: 0.16697423005986123 | D1-D2 diff: 2.113242372801436 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=80), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21760201801389445 Holdout data R^2 trained on entire dataset(80%): 0.14295021354684834 Dataset D1 R^2 on trained D1: 0.21711635783991579 .................................................. Pipeline #18: Score on D2: 0.16509200115820843 | D1-D2 diff: 2.130268900332756 Pipeline steps: SelectPercentile(percentile=60), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=19, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17554043023053922 Holdout data R^2 trained on entire dataset(80%): 0.1096481892320823 Dataset D1 R^2 on trained D1: 0.21365016893012845 .................................................. Pipeline #19: Score on D2: 0.16311978103763647 | D1-D2 diff: 2.151406599421793 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), SelectPercentile(percentile=80), SelectPercentile(percentile=95), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20164932505284583 Holdout data R^2 trained on entire dataset(80%): 0.14007399630788386 Dataset D1 R^2 on trained D1: 0.20979754182911747 .................................................. Pipeline #20: Score on D2: 0.15392919055151943 | D1-D2 diff: 2.2567277172872178 Pipeline steps: SelectPercentile(percentile=60), FeatureEncodingFrequencySelector(threshold=0.25), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=19, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15675917769313852 Holdout data R^2 trained on entire dataset(80%): 0.10164421888214892 Dataset D1 R^2 on trained D1: 0.19248442497949503 .................................................. Pipeline #21: Score on D2: 0.14231295957903056 | D1-D2 diff: 3.4784817699452333 Pipeline steps: OverDominanceEncoder(), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1478518176328717 Holdout data R^2 trained on entire dataset(80%): 0.14600056332102418 Dataset D1 R^2 on trained D1: 0.14914327993437415 .................................................. Pipeline #22: Score on D2: 0.14189296490617287 | D1-D2 diff: 3.492733958808586 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14699092019636806 Holdout data R^2 trained on entire dataset(80%): 0.14988763781105086 Dataset D1 R^2 on trained D1: 0.14861248060744248 .................................................. Pipeline #23: Score on D2: 0.12822759428885966 | D1-D2 diff: 3.5924828713331913 Pipeline steps: SelectPercentile(percentile=65), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=19, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12622157964825464 Holdout data R^2 trained on entire dataset(80%): 0.12603535942775212 Dataset D1 R^2 on trained D1: 0.13423132460098075 .................................................. Pipeline #24: Score on D2: 0.11831431643536117 | D1-D2 diff: 9.713096139020509 Pipeline steps: VarianceThreshold(threshold=0.25), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11951036220180511 Holdout data R^2 trained on entire dataset(80%): 0.12224411978633354 Dataset D1 R^2 on trained D1: 0.11842666544555669 .................................................. Pipeline #25: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.05), DominantEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 28 - 7 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.18566 Best score on D1-D2 diff: 6.83075 Gen 2 - Best score on D2: 0.19536 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.19536 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.19536 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.19536 Best score on D1-D2 diff: 11.96036 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: 0.0019254900725999002 Entire dataset(80%) R^2 trained on data (80%): 0.006832524901886905 Holdout R^2 (20%) trained on data (80%): 0.0032733862410158077 Dataset D1 R^2 on trained D1: 0.005335186881529896 Combined Dataset (100%) R^2 trained on combined data (100%): 0.006994559127800559 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.19536161188560552 | D1-D2 diff: 1.380413563127251 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.5, min_samples_leaf=2, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.47844608426370727 Holdout data R^2 trained on entire dataset(80%): 0.19399983588425185 Dataset D1 R^2 on trained D1: 0.47076110016351935 .................................................. Pipeline #2: Score on D2: 0.18966713446517747 | D1-D2 diff: 1.4336505940602382 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=3, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.43228624912221736 Holdout data R^2 trained on entire dataset(80%): 0.19223345372921563 Dataset D1 R^2 on trained D1: 0.4263826496020887 .................................................. Pipeline #3: Score on D2: 0.18920239843183684 | D1-D2 diff: 1.4381657685484766 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=3, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4310089868600935 Holdout data R^2 trained on entire dataset(80%): 0.1865694164077657 Dataset D1 R^2 on trained D1: 0.4229591753106524 .................................................. Pipeline #4: Score on D2: 0.18636919840840027 | D1-D2 diff: 1.5731758891509868 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3630996002481388 Holdout data R^2 trained on entire dataset(80%): 0.18570594594040024 Dataset D1 R^2 on trained D1: 0.3496333640552288 .................................................. Pipeline #5: Score on D2: 0.18577462364112896 | D1-D2 diff: 1.6840504466392439 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.4, min_samples_leaf=15, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3261138176326632 Holdout data R^2 trained on entire dataset(80%): 0.18610105773819596 Dataset D1 R^2 on trained D1: 0.31010567681829715 .................................................. Pipeline #6: Score on D2: 0.18265717008070526 | D1-D2 diff: 1.690166661254207 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.55, min_samples_leaf=17, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3220588807468886 Holdout data R^2 trained on entire dataset(80%): 0.18455393271937126 Dataset D1 R^2 on trained D1: 0.3051982987879601 .................................................. Pipeline #7: Score on D2: 0.18265374640291798 | D1-D2 diff: 1.7288540439794546 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.4, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.310396783065373 Holdout data R^2 trained on entire dataset(80%): 0.18563097113866078 Dataset D1 R^2 on trained D1: 0.29458894637508604 .................................................. Pipeline #8: Score on D2: 0.17932319097991722 | D1-D2 diff: 1.7891702314824878 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2922233453504063 Holdout data R^2 trained on entire dataset(80%): 0.18150991734429445 Dataset D1 R^2 on trained D1: 0.2769105006214081 .................................................. Pipeline #9: Score on D2: 0.1766735690427783 | D1-D2 diff: 1.7994427062481166 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=14, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2957195124913027 Holdout data R^2 trained on entire dataset(80%): 0.18512447568054002 Dataset D1 R^2 on trained D1: 0.27205150206771067 .................................................. Pipeline #10: Score on D2: 0.17620418868172405 | D1-D2 diff: 1.8505987946300704 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28228712975196746 Holdout data R^2 trained on entire dataset(80%): 0.1834561161788416 Dataset D1 R^2 on trained D1: 0.2614652825468333 .................................................. Pipeline #11: Score on D2: 0.1741166705077638 | D1-D2 diff: 1.8714108903948943 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2785293124698992 Holdout data R^2 trained on entire dataset(80%): 0.1842221970479434 Dataset D1 R^2 on trained D1: 0.2556477871014564 .................................................. Pipeline #12: Score on D2: 0.17012406630500299 | D1-D2 diff: 1.918303606703203 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.4, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2678922746119442 Holdout data R^2 trained on entire dataset(80%): 0.17788976453446992 Dataset D1 R^2 on trained D1: 0.24397068706150793 .................................................. Pipeline #13: Score on D2: 0.168870045767202 | D1-D2 diff: 1.9379595125000346 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2654472986006705 Holdout data R^2 trained on entire dataset(80%): 0.18308052723689416 Dataset D1 R^2 on trained D1: 0.2397659590653478 .................................................. Pipeline #14: Score on D2: 0.16576186227568446 | D1-D2 diff: 1.9553880761067493 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25278104292977455 Holdout data R^2 trained on entire dataset(80%): 0.17613447340641497 Dataset D1 R^2 on trained D1: 0.23416375984058946 .................................................. Pipeline #15: Score on D2: 0.13146736261143543 | D1-D2 diff: 1.9924878118621265 Pipeline steps: OverDominanceEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=15, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21176845016777368 Holdout data R^2 trained on entire dataset(80%): 0.14636806795295454 Dataset D1 R^2 on trained D1: 0.194915270458224 .................................................. Pipeline #16: Score on D2: 0.1314136106276117 | D1-D2 diff: 1.995613462707686 Pipeline steps: OverDominanceEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=15, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21031486891597861 Holdout data R^2 trained on entire dataset(80%): 0.14077215481313954 Dataset D1 R^2 on trained D1: 0.19446494754524413 .................................................. Pipeline #17: Score on D2: 0.122823737087707 | D1-D2 diff: 2.474960725835427 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=2, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1414137959002234 Holdout data R^2 trained on entire dataset(80%): 0.15006816976341764 Dataset D1 R^2 on trained D1: 0.14947554983408162 .................................................. Pipeline #18: Score on D2: 0.0996719614545577 | D1-D2 diff: 4.470964444933516 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=20, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10065464711395133 Holdout data R^2 trained on entire dataset(80%): 0.10507699510553048 Dataset D1 R^2 on trained D1: 0.10217458274714575 .................................................. Pipeline #19: Score on D2: 0.09541241223340657 | D1-D2 diff: 6.151473134315687 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09646696494666362 Holdout data R^2 trained on entire dataset(80%): 0.10383241065747018 Dataset D1 R^2 on trained D1: 0.09471404561280472 .................................................. Pipeline #20: Score on D2: 0.09541241223340646 | D1-D2 diff: 6.151473134315931 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09646696494666362 Holdout data R^2 trained on entire dataset(80%): 0.1038324106574704 Dataset D1 R^2 on trained D1: 0.09471404561280472 .................................................. Pipeline #21: Score on D2: 0.011635780430118259 | D1-D2 diff: 6.661600139970885 Pipeline steps: DominantEncoder(), HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011923788807055247 Holdout data R^2 trained on entire dataset(80%): 0.011586973860361227 Dataset D1 R^2 on trained D1: 0.012143572315990725 .................................................. Pipeline #22: Score on D2: 0.003934382697617256 | D1-D2 diff: 6.830747118156113 Pipeline steps: SelectPercentile(percentile=25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00538617304679434 Holdout data R^2 trained on entire dataset(80%): 0.005169386939627207 Dataset D1 R^2 on trained D1: 0.0034750502031356545 .................................................. Pipeline #23: Score on D2: 0.0021294252885081244 | D1-D2 diff: 7.882659740447309 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00228375837862127 Holdout data R^2 trained on entire dataset(80%): 0.00029050533806096457 Dataset D1 R^2 on trained D1: 0.0018704198529388982 .................................................. Pipeline #24: Score on D2: 0.0020727000823357322 | D1-D2 diff: 8.280150403883527 Pipeline steps: SelectPercentile(percentile=65), UnderDominanceEncoder(), VarianceThreshold(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0022250500274940688 Holdout data R^2 trained on entire dataset(80%): 0.0009772043141048625 Dataset D1 R^2 on trained D1: 0.0018599610900122965 .................................................. Pipeline #25: Score on D2: -4.886780840163141e-05 | D1-D2 diff: 11.960360895673112 Pipeline steps: DominantEncoder(), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0 Holdout data R^2 trained on entire dataset(80%): -0.0012028184006112053 Dataset D1 R^2 on trained D1: 0.0 .................................................. ************************************************************************************** Random Seed 28 - 8 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.17809 Best score on D1-D2 diff: 5.97130 Gen 2 - Best score on D2: 0.17928 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.17928 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.18123 Best score on D1-D2 diff: 11.96036 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: 0.0037679082865971214 Entire dataset(80%) R^2 trained on data (80%): 0.0065358975660850804 Holdout R^2 (20%) trained on data (80%): 0.005560106218087002 Dataset D1 R^2 on trained D1: 0.002926208175775824 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0071975933471795095 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.18123347028909753 | D1-D2 diff: 1.3703796295223902 Pipeline steps: OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.4, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.47369993267016886 Holdout data R^2 trained on entire dataset(80%): 0.20108659570118093 Dataset D1 R^2 on trained D1: 0.4647878912479676 .................................................. Pipeline #2: Score on D2: 0.1792813320043215 | D1-D2 diff: 1.559119977171886 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=6, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3582285584430088 Holdout data R^2 trained on entire dataset(80%): 0.20269280322697858 Dataset D1 R^2 on trained D1: 0.3485130857442843 .................................................. Pipeline #3: Score on D2: 0.17626701153934698 | D1-D2 diff: 1.7648414697145938 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29124464587664356 Holdout data R^2 trained on entire dataset(80%): 0.1987353782360347 Dataset D1 R^2 on trained D1: 0.27934767364772306 .................................................. Pipeline #4: Score on D2: 0.17281383594275057 | D1-D2 diff: 1.8806492238687977 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27219006193444795 Holdout data R^2 trained on entire dataset(80%): 0.19403399750634864 Dataset D1 R^2 on trained D1: 0.25275469365593795 .................................................. Pipeline #5: Score on D2: 0.14482385217168725 | D1-D2 diff: 1.8840059870597996 Pipeline steps: UnderDominanceEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=6, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1967658699752053 Holdout data R^2 trained on entire dataset(80%): 0.12186985679692275 Dataset D1 R^2 on trained D1: 0.22419650316627926 .................................................. Pipeline #6: Score on D2: 0.14252981182520896 | D1-D2 diff: 1.8867544550303 Pipeline steps: UnderDominanceEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=6, min_samples_leaf=16, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1962157956416375 Holdout data R^2 trained on entire dataset(80%): 0.12533368347177243 Dataset D1 R^2 on trained D1: 0.22144097834870113 .................................................. Pipeline #7: Score on D2: 0.11926945326141114 | D1-D2 diff: 2.0674959414707925 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), DecisionTreeRegressor(max_depth=5, min_samples_leaf=15, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15823950978649814 Holdout data R^2 trained on entire dataset(80%): 0.13293048059559698 Dataset D1 R^2 on trained D1: 0.1739989352018797 .................................................. Pipeline #8: Score on D2: 0.11456846103889096 | D1-D2 diff: 2.489023142960688 Pipeline steps: SelectPercentile(percentile=50), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12552096695357673 Holdout data R^2 trained on entire dataset(80%): 0.09003768298632875 Dataset D1 R^2 on trained D1: 0.1406230521054478 .................................................. Pipeline #9: Score on D2: 0.09353120088800793 | D1-D2 diff: 5.7014565399742905 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09592637491879097 Holdout data R^2 trained on entire dataset(80%): 0.10454165692992679 Dataset D1 R^2 on trained D1: 0.09447756165410104 .................................................. Pipeline #10: Score on D2: 0.0037826846555277793 | D1-D2 diff: 5.843712887340991 Pipeline steps: VarianceThreshold(threshold=0.25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0065247354733574126 Holdout data R^2 trained on entire dataset(80%): 0.0055228274798223564 Dataset D1 R^2 on trained D1: 0.0029251641386491833 .................................................. Pipeline #11: Score on D2: 0.0037695551199434796 | D1-D2 diff: 5.867208154007252 Pipeline steps: VarianceThreshold(threshold=0.15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.006533018239813448 Holdout data R^2 trained on entire dataset(80%): 0.005608166923752322 Dataset D1 R^2 on trained D1: 0.0029256880977132083 .................................................. Pipeline #12: Score on D2: 0.0037679082865971214 | D1-D2 diff: 5.870980716372428 Pipeline steps: LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0065358975660850804 Holdout data R^2 trained on entire dataset(80%): 0.005560106218087002 Dataset D1 R^2 on trained D1: 0.002926208175775824 .................................................. Pipeline #13: Score on D2: 0.003644138211574166 | D1-D2 diff: 6.030122405479978 Pipeline steps: SelectPercentile(percentile=80), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.006522214189410791 Holdout data R^2 trained on entire dataset(80%): 0.0053527256372890575 Dataset D1 R^2 on trained D1: 0.002887835794961102 .................................................. Pipeline #14: Score on D2: 0.003629392884037963 | D1-D2 diff: 6.105263168237238 Pipeline steps: SelectPercentile(percentile=85), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.006530568076163812 Holdout data R^2 trained on entire dataset(80%): 0.005446272565252763 Dataset D1 R^2 on trained D1: 0.0029096416072462716 .................................................. Pipeline #15: Score on D2: 0.0007442725284411145 | D1-D2 diff: 8.392773755820683 Pipeline steps: SelectPercentile(percentile=5), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.003662721374289468 Holdout data R^2 trained on entire dataset(80%): 0.005466204609214653 Dataset D1 R^2 on trained D1: 0.0005427247868708962 .................................................. Pipeline #16: Score on D2: -4.886780840163141e-05 | D1-D2 diff: 11.960360895673112 Pipeline steps: HeterosisEncoder(), VarianceThreshold(), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0 Holdout data R^2 trained on entire dataset(80%): -0.0012028184006112053 Dataset D1 R^2 on trained D1: 0.0 .................................................. ************************************************************************************** Random Seed 28 - 9 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.18215 Best score on D1-D2 diff: 10.23196 Gen 2 - Best score on D2: 0.18393 Best score on D1-D2 diff: 10.39912 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: 0.011309597680826067 Entire dataset(80%) R^2 trained on data (80%): 0.012698819726665422 Holdout R^2 (20%) trained on data (80%): 0.00305038369119337 Dataset D1 R^2 on trained D1: 0.007847929334236725 Combined Dataset (100%) R^2 trained on combined data (100%): 0.011685413568843006 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1839253551431811 | D1-D2 diff: 1.4076165572570776 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.5, min_samples_leaf=7, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4370139053815145 Holdout data R^2 trained on entire dataset(80%): 0.21276938481541396 Dataset D1 R^2 on trained D1: 0.43864505450058966 .................................................. Pipeline #2: Score on D2: 0.18220690101259884 | D1-D2 diff: 1.6771779221270506 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=11, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.323650206734744 Holdout data R^2 trained on entire dataset(80%): 0.2173833410412066 Dataset D1 R^2 on trained D1: 0.308588385587784 .................................................. Pipeline #3: Score on D2: 0.18214973888191344 | D1-D2 diff: 1.7875182217621721 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29207262968619796 Holdout data R^2 trained on entire dataset(80%): 0.20893915393154916 Dataset D1 R^2 on trained D1: 0.2800983065189967 .................................................. Pipeline #4: Score on D2: 0.17815721854268207 | D1-D2 diff: 1.9003892206495405 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2701226751042135 Holdout data R^2 trained on entire dataset(80%): 0.20641062531158905 Dataset D1 R^2 on trained D1: 0.25482797825000225 .................................................. Pipeline #5: Score on D2: 0.12746615519590232 | D1-D2 diff: 2.0549961078871037 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=5, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17067905790424676 Holdout data R^2 trained on entire dataset(80%): 0.15073774717969812 Dataset D1 R^2 on trained D1: 0.18353943835591535 .................................................. Pipeline #6: Score on D2: 0.12534993127423966 | D1-D2 diff: 2.594791974878421 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1443265941828743 Holdout data R^2 trained on entire dataset(80%): 0.14913049333044848 Dataset D1 R^2 on trained D1: 0.14740913419095236 .................................................. Pipeline #7: Score on D2: 0.11244889255996027 | D1-D2 diff: 3.1062882646945256 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12417147653801519 Holdout data R^2 trained on entire dataset(80%): 0.12514480739421852 Dataset D1 R^2 on trained D1: 0.12318960219988406 .................................................. Pipeline #8: Score on D2: 0.10064053231378767 | D1-D2 diff: 3.5099961870536154 Pipeline steps: SelectPercentile(percentile=90), DecisionTreeRegressor(max_depth=3, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10260539494003573 Holdout data R^2 trained on entire dataset(80%): 0.10876135918352547 Dataset D1 R^2 on trained D1: 0.1072288331418858 .................................................. Pipeline #9: Score on D2: 0.09974126296093677 | D1-D2 diff: 4.40325885503275 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0988633792889485 Holdout data R^2 trained on entire dataset(80%): 0.09410056521969301 Dataset D1 R^2 on trained D1: 0.09708113130473617 .................................................. Pipeline #10: Score on D2: 0.054867269076522396 | D1-D2 diff: 5.03381160426639 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05409243485585624 Holdout data R^2 trained on entire dataset(80%): 0.057958880890399245 Dataset D1 R^2 on trained D1: 0.05330982604589052 .................................................. Pipeline #11: Score on D2: 0.051874568777007135 | D1-D2 diff: 5.385437902892513 Pipeline steps: SelectPercentile(percentile=10), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05213290906271395 Holdout data R^2 trained on entire dataset(80%): 0.052166128659852484 Dataset D1 R^2 on trained D1: 0.05068574930580794 .................................................. Pipeline #12: Score on D2: 0.005820894763397422 | D1-D2 diff: 10.399123843150548 Pipeline steps: SelectPercentile(percentile=20), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.010637885719036011 Holdout data R^2 trained on entire dataset(80%): 0.0027431512834453775 Dataset D1 R^2 on trained D1: 0.005735385532743997 .................................................. ************************************************************************************** Random Seed 29 - 0 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.00083 Best score on D1-D2 diff: 5.60117 Gen 2 - Best score on D2: 0.00083 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00209 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00257 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00257 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.00340 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.00398 Best score on D1-D2 diff: 14.53993 Gen 19 - Best score on D2: 0.00398 Best score on D1-D2 diff: 14.53993 Gen 20 - Best score on D2: 0.00398 Best score on D1-D2 diff: 14.53993 Gen 21 - Best score on D2: 0.00398 Best score on D1-D2 diff: 14.53993 Gen 22 - Best score on D2: 0.00398 Best score on D1-D2 diff: 14.53993 Gen 23 - Best score on D2: 0.00398 Best score on D1-D2 diff: 14.53993 Gen 24 - Best score on D2: 0.00398 Best score on D1-D2 diff: 14.53993 Gen 25 - Best score on D2: 0.00398 Best score on D1-D2 diff: 16.84840 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.01218966803570054 Entire dataset(80%) R^2 trained on data (80%): 0.004494251754839307 Holdout R^2 (20%) trained on data (80%): -0.007858066378064299 Dataset D1 R^2 on trained D1: 0.010680019843458788 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0028343599390574514 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.003983848287373193 | D1-D2 diff: 5.493216557748918 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RecessiveEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0038380976185211635 Holdout data R^2 trained on entire dataset(80%): -0.0030722525840156667 Dataset D1 R^2 on trained D1: 0.00288561875469151 .................................................. Pipeline #2: Score on D2: 0.001992184916600137 | D1-D2 diff: 16.84840249418592 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RecessiveEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0026151494762503003 Holdout data R^2 trained on entire dataset(80%): -0.005606788096129245 Dataset D1 R^2 on trained D1: 0.0020045947252319563 .................................................. ************************************************************************************** Random Seed 29 - 1 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.05509 Best score on D1-D2 diff: 5.60117 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.01470380495108281 Entire dataset(80%) R^2 trained on data (80%): 0.004419210679260432 Holdout R^2 (20%) trained on data (80%): -0.010074242827978974 Dataset D1 R^2 on trained D1: 0.011757585883501842 Combined Dataset (100%) R^2 trained on combined data (100%): 0.002523116691498961 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.05508595927028337 | D1-D2 diff: 1.5202426423064068 Pipeline steps: RandomForestRegressor(bootstrap=False, max_features=0.7000000000000001, min_samples_leaf=18, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21798456382783438 Holdout data R^2 trained on entire dataset(80%): 0.036281265815978014 Dataset D1 R^2 on trained D1: 0.24230428593920295 .................................................. Pipeline #2: Score on D2: 0.054973509074234594 | D1-D2 diff: 2.0368367265731644 Pipeline steps: VarianceThreshold(threshold=0.25), DecisionTreeRegressor(max_depth=5, min_samples_leaf=20, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0204893909324676 Holdout data R^2 trained on entire dataset(80%): -0.025744573269826487 Dataset D1 R^2 on trained D1: 0.11307337512269455 .................................................. Pipeline #3: Score on D2: 0.039878430073915117 | D1-D2 diff: 2.2893533707286795 Pipeline steps: DecisionTreeRegressor(max_depth=4, min_samples_leaf=10, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.012685730313234833 Holdout data R^2 trained on entire dataset(80%): -0.018008639658610992 Dataset D1 R^2 on trained D1: 0.07628239341300946 .................................................. Pipeline #4: Score on D2: 0.0008325348022591994 | D1-D2 diff: 5.601170657297559 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0014071918318177001 Holdout data R^2 trained on entire dataset(80%): -0.0037128949699982705 Dataset D1 R^2 on trained D1: 0.0018485139160190345 .................................................. ************************************************************************************** Random Seed 29 - 2 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.08299 Best score on D1-D2 diff: 3.81307 Gen 2 - Best score on D2: 0.09200 Best score on D1-D2 diff: 3.81307 Gen 3 - Best score on D2: 0.09813 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.09813 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.09813 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.10060 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.10060 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.10237 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.10237 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.10237 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.10444 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.10444 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.10444 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.10444 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.10444 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.10444 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.10444 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.10444 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.10444 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.10477 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.10477 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.10477 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.10477 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.10477 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.10477 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.012587377438291192 Entire dataset(80%) R^2 trained on data (80%): 0.0036942494696793338 Holdout R^2 (20%) trained on data (80%): -0.010981041451631368 Dataset D1 R^2 on trained D1: 0.010744656216632564 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0017125000541831081 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.10477139619274167 | D1-D2 diff: 1.6013260793300446 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.7500000000000001, min_samples_leaf=11, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22409763142923433 Holdout data R^2 trained on entire dataset(80%): 0.089340381218598 Dataset D1 R^2 on trained D1: 0.25685447410299667 .................................................. Pipeline #2: Score on D2: 0.10444361138033209 | D1-D2 diff: 1.7741509986697697 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.7500000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17861347823676743 Holdout data R^2 trained on entire dataset(80%): 0.09632203919473326 Dataset D1 R^2 on trained D1: 0.20537765675237463 .................................................. Pipeline #3: Score on D2: 0.10266011166502009 | D1-D2 diff: 1.819759858856224 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1673260147300456 Holdout data R^2 trained on entire dataset(80%): 0.0931730740869432 Dataset D1 R^2 on trained D1: 0.19384936849122525 .................................................. Pipeline #4: Score on D2: 0.09952200394533262 | D1-D2 diff: 1.8485804048328058 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15686060716959083 Holdout data R^2 trained on entire dataset(80%): 0.08729208270311695 Dataset D1 R^2 on trained D1: 0.1851560806118121 .................................................. Pipeline #5: Score on D2: 0.09224500218200526 | D1-D2 diff: 1.859048167559938 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=20, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18438285680472832 Holdout data R^2 trained on entire dataset(80%): 0.11907486806138279 Dataset D1 R^2 on trained D1: 0.1759665850004184 .................................................. Pipeline #6: Score on D2: 0.08582915774777711 | D1-D2 diff: 1.8763439720695874 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=18, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18256502832891275 Holdout data R^2 trained on entire dataset(80%): 0.11424036836695373 Dataset D1 R^2 on trained D1: 0.16650623838370404 .................................................. Pipeline #7: Score on D2: 0.08176009959494401 | D1-D2 diff: 1.911762621005272 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16660955239466368 Holdout data R^2 trained on entire dataset(80%): 0.1045910043088174 Dataset D1 R^2 on trained D1: 0.15662256684113907 .................................................. Pipeline #8: Score on D2: 0.07924596503749326 | D1-D2 diff: 1.9467511332281415 Pipeline steps: VarianceThreshold(threshold=0.05), HeterosisEncoder(), VarianceThreshold(threshold=0.2), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16106674338423732 Holdout data R^2 trained on entire dataset(80%): 0.10374933084951699 Dataset D1 R^2 on trained D1: 0.14886985037691824 .................................................. Pipeline #9: Score on D2: 0.07652042812340731 | D1-D2 diff: 2.288760129146611 Pipeline steps: VarianceThreshold(threshold=0.3), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=20, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05768788228798327 Holdout data R^2 trained on entire dataset(80%): 0.0424024103709294 Dataset D1 R^2 on trained D1: 0.11296214944782867 .................................................. Pipeline #10: Score on D2: 0.07163618187137255 | D1-D2 diff: 2.5709375513628334 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), SelectPercentile(percentile=90), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.024920658370945925 Holdout data R^2 trained on entire dataset(80%): 0.0035480377427561383 Dataset D1 R^2 on trained D1: 0.09452555450124123 .................................................. Pipeline #11: Score on D2: 0.05345239337498042 | D1-D2 diff: 2.673738689661639 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=90), VarianceThreshold(threshold=0.15), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.013304936974514403 Holdout data R^2 trained on entire dataset(80%): -0.010974742206024901 Dataset D1 R^2 on trained D1: 0.07301938946405306 .................................................. Pipeline #12: Score on D2: 0.028952471114362943 | D1-D2 diff: 3.0078750407422534 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), SelectPercentile(percentile=90), DecisionTreeRegressor(max_depth=3, min_samples_leaf=13, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.006365411850871805 Holdout data R^2 trained on entire dataset(80%): -0.005317264035376024 Dataset D1 R^2 on trained D1: 0.041169366083887216 .................................................. Pipeline #13: Score on D2: 1.855639124492825e-05 | D1-D2 diff: 5.495481164941277 Pipeline steps: SelectPercentile(percentile=5), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0005595072635027343 Holdout data R^2 trained on entire dataset(80%): -0.0012805403676909854 Dataset D1 R^2 on trained D1: 0.0011149767854539139 .................................................. Pipeline #14: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: DominantEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 29 - 3 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.11136 Best score on D1-D2 diff: 7.11944 Gen 2 - Best score on D2: 0.11819 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.11819 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.11819 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.11850 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.12689 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.12890 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.12890 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.12890 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.12890 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.12890 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.12890 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.12890 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.12890 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.12890 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.008261003121336152 Entire dataset(80%) R^2 trained on data (80%): 0.005557227217408811 Holdout R^2 (20%) trained on data (80%): -0.011383796347386621 Dataset D1 R^2 on trained D1: 0.009053912984209211 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003043530652387494 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.12890282659555852 | D1-D2 diff: 1.581219328985427 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.55, min_samples_leaf=7, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.289892384216076 Holdout data R^2 trained on entire dataset(80%): 0.13785453033581452 Dataset D1 R^2 on trained D1: 0.2888702470781205 .................................................. Pipeline #2: Score on D2: 0.12708214486039282 | D1-D2 diff: 1.6916131059665165 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), FeatureEncodingFrequencySelector(threshold=0.0), RandomForestRegressor(max_features=0.55, min_samples_leaf=12, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24716449527080608 Holdout data R^2 trained on entire dataset(80%): 0.13644365424558136 Dataset D1 R^2 on trained D1: 0.24920468670131113 .................................................. Pipeline #3: Score on D2: 0.12656937421465797 | D1-D2 diff: 1.7781083768024468 Pipeline steps: VarianceThreshold(threshold=0.1), HeterosisEncoder(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22596731658184355 Holdout data R^2 trained on entire dataset(80%): 0.1320834888740341 Dataset D1 R^2 on trained D1: 0.22660785508958925 .................................................. Pipeline #4: Score on D2: 0.1264167354094926 | D1-D2 diff: 1.8402565216150721 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.55, min_samples_leaf=17, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21664173706042156 Holdout data R^2 trained on entire dataset(80%): 0.13490277474136692 Dataset D1 R^2 on trained D1: 0.21361072272052917 .................................................. Pipeline #5: Score on D2: 0.12281240826640671 | D1-D2 diff: 1.9214731205193254 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), VarianceThreshold(threshold=0.35), RandomForestRegressor(max_features=0.45, min_samples_leaf=20, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2004737548428448 Holdout data R^2 trained on entire dataset(80%): 0.12886973396239 Dataset D1 R^2 on trained D1: 0.19617298653162418 .................................................. Pipeline #6: Score on D2: 0.11187238357137819 | D1-D2 diff: 2.010360421042721 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), VarianceThreshold(threshold=0.35), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17923853984422633 Holdout data R^2 trained on entire dataset(80%): 0.11888647432404365 Dataset D1 R^2 on trained D1: 0.17309393035546983 .................................................. Pipeline #7: Score on D2: 0.09757786753468467 | D1-D2 diff: 2.021047302683627 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=80), FeatureEncodingFrequencySelector(threshold=0.35), VarianceThreshold(threshold=0.2), RandomForestRegressor(max_features=0.3, min_samples_leaf=15, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1748137186167582 Holdout data R^2 trained on entire dataset(80%): 0.1071969854703172 Dataset D1 R^2 on trained D1: 0.15751474127591758 .................................................. Pipeline #8: Score on D2: 0.09611873513011548 | D1-D2 diff: 2.0761610616122876 Pipeline steps: VarianceThreshold(threshold=0.2), HeterosisEncoder(), SelectPercentile(percentile=80), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15757898543793314 Holdout data R^2 trained on entire dataset(80%): 0.10258727479068541 Dataset D1 R^2 on trained D1: 0.14994023962717906 .................................................. Pipeline #9: Score on D2: 0.08838938550664888 | D1-D2 diff: 2.938228077898735 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09540464654092251 Holdout data R^2 trained on entire dataset(80%): 0.08372645743322937 Dataset D1 R^2 on trained D1: 0.10180646465529597 .................................................. Pipeline #10: Score on D2: 0.015614169192261662 | D1-D2 diff: 3.899716634664504 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09540464654092251 Holdout data R^2 trained on entire dataset(80%): 0.08372645743322937 Dataset D1 R^2 on trained D1: 0.01993799108317018 .................................................. Pipeline #11: Score on D2: 0.001998417583817247 | D1-D2 diff: 7.197750800452036 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0021847691113058287 Holdout data R^2 trained on entire dataset(80%): -0.0023728607196011886 Dataset D1 R^2 on trained D1: 0.0023709917807550607 .................................................. Pipeline #12: Score on D2: 2.4123932358399713e-05 | D1-D2 diff: 11.376247970679298 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=5), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00026305671088988625 Holdout data R^2 trained on entire dataset(80%): -0.0011276262548705063 Dataset D1 R^2 on trained D1: 8.382798346662224e-05 .................................................. Pipeline #13: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: RecessiveEncoder(), OverDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=12, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 29 - 4 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.12755 Best score on D1-D2 diff: 3.72306 Gen 2 - Best score on D2: 0.12849 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.13438 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.13939 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.13939 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.13939 Best score on D1-D2 diff: 11.96036 Gen 7 - Best score on D2: 0.13946 Best score on D1-D2 diff: 11.96036 Gen 8 - Best score on D2: 0.13946 Best score on D1-D2 diff: 11.96036 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.011710635112184509 Entire dataset(80%) R^2 trained on data (80%): 0.0060272895534457804 Holdout R^2 (20%) trained on data (80%): -0.011775214356742003 Dataset D1 R^2 on trained D1: 0.011658762676047374 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003360050139953552 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1394559065680967 | D1-D2 diff: 1.6337440338067772 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.5, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28550743731470296 Holdout data R^2 trained on entire dataset(80%): 0.1694923700650962 Dataset D1 R^2 on trained D1: 0.27982255864265737 .................................................. Pipeline #2: Score on D2: 0.1393873898966561 | D1-D2 diff: 1.7185636320251239 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=11, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2559679013023348 Holdout data R^2 trained on entire dataset(80%): 0.16759316521347578 Dataset D1 R^2 on trained D1: 0.2540277473852274 .................................................. Pipeline #3: Score on D2: 0.13834013645034593 | D1-D2 diff: 1.8068910445319672 Pipeline steps: SelectPercentile(percentile=90), HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2330449410989356 Holdout data R^2 trained on entire dataset(80%): 0.16477056475491136 Dataset D1 R^2 on trained D1: 0.23215510522810057 .................................................. Pipeline #4: Score on D2: 0.13828887006484114 | D1-D2 diff: 1.80814151230094 Pipeline steps: SelectPercentile(percentile=90), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2346312656106202 Holdout data R^2 trained on entire dataset(80%): 0.16581206346423272 Dataset D1 R^2 on trained D1: 0.231844587110153 .................................................. Pipeline #5: Score on D2: 0.13346405473518397 | D1-D2 diff: 1.8158183568678599 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22951307260648046 Holdout data R^2 trained on entire dataset(80%): 0.1555879383541552 Dataset D1 R^2 on trained D1: 0.2254476522509059 .................................................. Pipeline #6: Score on D2: 0.12546092779904494 | D1-D2 diff: 1.8469563938274465 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2144150437919794 Holdout data R^2 trained on entire dataset(80%): 0.14571108971905722 Dataset D1 R^2 on trained D1: 0.21139659082939133 .................................................. Pipeline #7: Score on D2: 0.1217249211908985 | D1-D2 diff: 1.9028837299303023 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20488488913701453 Holdout data R^2 trained on entire dataset(80%): 0.14901111492543306 Dataset D1 R^2 on trained D1: 0.19799443689519858 .................................................. Pipeline #8: Score on D2: 0.10578910623013305 | D1-D2 diff: 2.3720443280538115 Pipeline steps: VarianceThreshold(threshold=0.1), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), DecisionTreeRegressor(max_depth=4, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12606050246457834 Holdout data R^2 trained on entire dataset(80%): 0.10232869045229698 Dataset D1 R^2 on trained D1: 0.13737613653463 .................................................. Pipeline #9: Score on D2: 0.10152967473515406 | D1-D2 diff: 2.378167800297737 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12606050246457834 Holdout data R^2 trained on entire dataset(80%): 0.10232869045229698 Dataset D1 R^2 on trained D1: 0.13279262945672088 .................................................. Pipeline #10: Score on D2: 0.08999015437281754 | D1-D2 diff: 4.000285907331138 Pipeline steps: SelectPercentile(percentile=15), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.5, min_samples_leaf=16, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0023743970029727057 Holdout data R^2 trained on entire dataset(80%): -0.0009278066367828242 Dataset D1 R^2 on trained D1: 0.09389528774684464 .................................................. Pipeline #11: Score on D2: 0.0015878911245200689 | D1-D2 diff: 5.214195981794383 Pipeline steps: SelectPercentile(percentile=55), UnderDominanceEncoder(), SelectPercentile(percentile=15), SelectPercentile(percentile=15), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.5, min_samples_leaf=16, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00010114502442692963 Holdout data R^2 trained on entire dataset(80%): -0.0020176497803925386 Dataset D1 R^2 on trained D1: 0.00023503816038461522 .................................................. Pipeline #12: Score on D2: -4.886780840163141e-05 | D1-D2 diff: 11.960360895673112 Pipeline steps: HeterosisEncoder(), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0 Holdout data R^2 trained on entire dataset(80%): -0.0012028184006112053 Dataset D1 R^2 on trained D1: 0.0 .................................................. ************************************************************************************** Random Seed 29 - 5 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.14603 Best score on D1-D2 diff: 5.60238 Gen 2 - Best score on D2: 0.15440 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.15440 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15440 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.15541 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.15541 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15541 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.15775 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.16010 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.16229 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.16229 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.16229 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.16229 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.16229 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.16229 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.16229 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.16295 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.16295 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.16295 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.16295 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.16295 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.16295 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.16319 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.16319 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.16319 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.009074890692971627 Entire dataset(80%) R^2 trained on data (80%): 0.0059530626408842435 Holdout R^2 (20%) trained on data (80%): -0.009811632281629068 Dataset D1 R^2 on trained D1: 0.011005211698654094 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003854025703159958 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1631858793043257 | D1-D2 diff: 1.5800470001919247 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.55, min_samples_leaf=6, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3319507181828262 Holdout data R^2 trained on entire dataset(80%): 0.1739306122740426 Dataset D1 R^2 on trained D1: 0.3236285849659376 .................................................. Pipeline #2: Score on D2: 0.15937549854824995 | D1-D2 diff: 1.5806778177350496 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.55, min_samples_leaf=6, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33176026483135357 Holdout data R^2 trained on entire dataset(80%): 0.17095522427103593 Dataset D1 R^2 on trained D1: 0.31956223932708827 .................................................. Pipeline #3: Score on D2: 0.1577468641105445 | D1-D2 diff: 1.6715427865603496 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=11, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2942833774887207 Holdout data R^2 trained on entire dataset(80%): 0.16384448177772182 Dataset D1 R^2 on trained D1: 0.2858412243844961 .................................................. Pipeline #4: Score on D2: 0.15566629212567262 | D1-D2 diff: 1.8310534598284582 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), UnderDominanceEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2552760656517663 Holdout data R^2 trained on entire dataset(80%): 0.1634757192497318 Dataset D1 R^2 on trained D1: 0.2446265233463515 .................................................. Pipeline #5: Score on D2: 0.1547262308445566 | D1-D2 diff: 1.9222511713294301 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RecessiveEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=19, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24046666603013744 Holdout data R^2 trained on entire dataset(80%): 0.16413792811996653 Dataset D1 R^2 on trained D1: 0.22796810742700147 .................................................. Pipeline #6: Score on D2: 0.14857880025787673 | D1-D2 diff: 1.9469595270366402 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=20, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23380136383054373 Holdout data R^2 trained on entire dataset(80%): 0.1599938828390597 Dataset D1 R^2 on trained D1: 0.21817288147020852 .................................................. Pipeline #7: Score on D2: 0.1478107067187493 | D1-D2 diff: 1.9596345425297896 Pipeline steps: VarianceThreshold(threshold=0.05), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=20, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2276819562713176 Holdout data R^2 trained on entire dataset(80%): 0.16040329704011036 Dataset D1 R^2 on trained D1: 0.2156216296483071 .................................................. Pipeline #8: Score on D2: 0.1391204403585622 | D1-D2 diff: 2.0234698326070886 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2102476327573255 Holdout data R^2 trained on entire dataset(80%): 0.1573569281554168 Dataset D1 R^2 on trained D1: 0.19877079966934874 .................................................. Pipeline #9: Score on D2: 0.13433664730452743 | D1-D2 diff: 2.0401761404632253 Pipeline steps: VarianceThreshold(threshold=0.05), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), SelectPercentile(percentile=65), RandomForestRegressor(max_features=0.55, min_samples_leaf=19, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17576283148997562 Holdout data R^2 trained on entire dataset(80%): 0.12897242357668282 Dataset D1 R^2 on trained D1: 0.19205704875453544 .................................................. Pipeline #10: Score on D2: 0.1305114150033886 | D1-D2 diff: 2.0977943210962513 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19782577957376313 Holdout data R^2 trained on entire dataset(80%): 0.14705410383517759 Dataset D1 R^2 on trained D1: 0.1821469140062032 .................................................. Pipeline #11: Score on D2: 0.1231722030473088 | D1-D2 diff: 2.2044175996709567 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=80), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16922858511744965 Holdout data R^2 trained on entire dataset(80%): 0.13056255914887271 Dataset D1 R^2 on trained D1: 0.1655193856113324 .................................................. Pipeline #12: Score on D2: 0.11345594537426529 | D1-D2 diff: 2.211642338745153 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16662094855614085 Holdout data R^2 trained on entire dataset(80%): 0.1361117167025614 Dataset D1 R^2 on trained D1: 0.15525249377662576 .................................................. Pipeline #13: Score on D2: 0.10403600820800274 | D1-D2 diff: 2.632898821329631 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0938239235023014 Holdout data R^2 trained on entire dataset(80%): 0.0972249322057156 Dataset D1 R^2 on trained D1: 0.0832264279820567 .................................................. Pipeline #14: Score on D2: 0.10403600820800263 | D1-D2 diff: 2.6328988213296345 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=80), OverDominanceEncoder(), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0938239235023014 Holdout data R^2 trained on entire dataset(80%): 0.09722493220571538 Dataset D1 R^2 on trained D1: 0.0832264279820567 .................................................. Pipeline #15: Score on D2: 0.10057281330179912 | D1-D2 diff: 2.8615229033548255 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), DecisionTreeRegressor(max_depth=3, min_samples_leaf=16, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10260545893285755 Holdout data R^2 trained on entire dataset(80%): 0.1267262535671162 Dataset D1 R^2 on trained D1: 0.11548739559521892 .................................................. Pipeline #16: Score on D2: 0.09893306385090284 | D1-D2 diff: 3.4737083556485477 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), HeterosisEncoder(), RecessiveEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=16, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09832921026767838 Holdout data R^2 trained on entire dataset(80%): 0.08775919431771917 Dataset D1 R^2 on trained D1: 0.09206512235825615 .................................................. Pipeline #17: Score on D2: 0.09893306385090272 | D1-D2 diff: 3.473708355648562 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09832921026767838 Holdout data R^2 trained on entire dataset(80%): 0.08775919431771917 Dataset D1 R^2 on trained D1: 0.09206512235825615 .................................................. Pipeline #18: Score on D2: 0.09585372984019302 | D1-D2 diff: 5.1307230513235265 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=75), VarianceThreshold(threshold=0.15), DecisionTreeRegressor(max_depth=3, min_samples_leaf=4, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09880923150358811 Holdout data R^2 trained on entire dataset(80%): 0.0883206476505567 Dataset D1 R^2 on trained D1: 0.09441066545284849 .................................................. Pipeline #19: Score on D2: 0.09513425897672545 | D1-D2 diff: 5.324187084417244 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=16, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09926567060623404 Holdout data R^2 trained on entire dataset(80%): 0.08613065692504107 Dataset D1 R^2 on trained D1: 0.09388978227702816 .................................................. Pipeline #20: Score on D2: 0.09359313862268648 | D1-D2 diff: 8.753344165290185 Pipeline steps: OverDominanceEncoder(), VarianceThreshold(threshold=0.25), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09900746267824512 Holdout data R^2 trained on entire dataset(80%): 0.08837253522191213 Dataset D1 R^2 on trained D1: 0.0937634736569326 .................................................. Pipeline #21: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 29 - 6 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.16003 Best score on D1-D2 diff: 5.60238 Gen 2 - Best score on D2: 0.16357 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.16919 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.16919 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.16919 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.16919 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.16919 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.16919 Best score on D1-D2 diff: 15.31265 Gen 9 - Best score on D2: 0.16919 Best score on D1-D2 diff: 15.31265 Gen 10 - Best score on D2: 0.16919 Best score on D1-D2 diff: 15.31265 Gen 11 - Best score on D2: 0.16919 Best score on D1-D2 diff: 15.31265 Gen 12 - Best score on D2: 0.16919 Best score on D1-D2 diff: 15.31265 Gen 13 - Best score on D2: 0.16937 Best score on D1-D2 diff: 15.31265 Gen 14 - Best score on D2: 0.16937 Best score on D1-D2 diff: 23.56615 Gen 15 - Best score on D2: 0.16937 Best score on D1-D2 diff: 23.56615 Gen 16 - Best score on D2: 0.16937 Best score on D1-D2 diff: 23.56615 Gen 17 - Best score on D2: 0.16937 Best score on D1-D2 diff: 23.56615 Gen 18 - Best score on D2: 0.16937 Best score on D1-D2 diff: 23.56615 Gen 19 - Best score on D2: 0.16937 Best score on D1-D2 diff: 23.56615 Gen 20 - Best score on D2: 0.16937 Best score on D1-D2 diff: 23.56615 Gen 21 - Best score on D2: 0.16937 Best score on D1-D2 diff: 23.56615 Gen 22 - Best score on D2: 0.16937 Best score on D1-D2 diff: 23.56615 Gen 23 - Best score on D2: 0.16937 Best score on D1-D2 diff: 23.56615 Gen 24 - Best score on D2: 0.17272 Best score on D1-D2 diff: 23.56615 Gen 25 - Best score on D2: 0.17272 Best score on D1-D2 diff: 23.56615 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.009566446845737486 Entire dataset(80%) R^2 trained on data (80%): 0.005949795183004425 Holdout R^2 (20%) trained on data (80%): -0.008307790198895049 Dataset D1 R^2 on trained D1: 0.01121567513429278 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004125836578377062 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1727206985255424 | D1-D2 diff: 1.5633082330168164 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=3, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33919854905772684 Holdout data R^2 trained on entire dataset(80%): 0.20025741557035637 Dataset D1 R^2 on trained D1: 0.3401461735276058 .................................................. Pipeline #2: Score on D2: 0.1681538627313559 | D1-D2 diff: 1.627707859721647 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=10, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31542362537533075 Holdout data R^2 trained on entire dataset(80%): 0.19184045458380627 Dataset D1 R^2 on trained D1: 0.3106142621758864 .................................................. Pipeline #3: Score on D2: 0.16528682170924713 | D1-D2 diff: 1.8095007371241043 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=95), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2541539522785946 Holdout data R^2 trained on entire dataset(80%): 0.1774959409622746 Dataset D1 R^2 on trained D1: 0.25856175402286286 .................................................. Pipeline #4: Score on D2: 0.1629688490485015 | D1-D2 diff: 1.8200446027962995 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=95), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25626661883775326 Holdout data R^2 trained on entire dataset(80%): 0.17876706212312965 Dataset D1 R^2 on trained D1: 0.2541010534366597 .................................................. Pipeline #5: Score on D2: 0.16165476228499465 | D1-D2 diff: 1.9256269915451008 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=95), HeterosisEncoder(), VarianceThreshold(), RandomForestRegressor(max_features=0.45, min_samples_leaf=18, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2358662575189736 Holdout data R^2 trained on entire dataset(80%): 0.17568124866495305 Dataset D1 R^2 on trained D1: 0.23438438601078704 .................................................. Pipeline #6: Score on D2: 0.15882166072708293 | D1-D2 diff: 1.9463926580548707 Pipeline steps: VarianceThreshold(), HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23639380894130935 Holdout data R^2 trained on entire dataset(80%): 0.17848720452074707 Dataset D1 R^2 on trained D1: 0.22849685191201863 .................................................. Pipeline #7: Score on D2: 0.1498700142995908 | D1-D2 diff: 2.0012434762969202 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22253317264951678 Holdout data R^2 trained on entire dataset(80%): 0.17352149815078188 Dataset D1 R^2 on trained D1: 0.21221482106133305 .................................................. Pipeline #8: Score on D2: 0.1444376004972482 | D1-D2 diff: 2.0821364399126727 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21154197112908335 Holdout data R^2 trained on entire dataset(80%): 0.16475619784796525 Dataset D1 R^2 on trained D1: 0.19764392518570872 .................................................. Pipeline #9: Score on D2: 0.14031019389366783 | D1-D2 diff: 2.101473719865062 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20663079339476387 Holdout data R^2 trained on entire dataset(80%): 0.1630804027020164 Dataset D1 R^2 on trained D1: 0.19158501418302698 .................................................. Pipeline #10: Score on D2: 0.13410490634093464 | D1-D2 diff: 2.1315615682019904 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1967269337245945 Holdout data R^2 trained on entire dataset(80%): 0.16047875239786558 Dataset D1 R^2 on trained D1: 0.18254539042378226 .................................................. Pipeline #11: Score on D2: 0.11974907195102513 | D1-D2 diff: 2.5231125002510417 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1427763425017199 Holdout data R^2 trained on entire dataset(80%): 0.13287537059739996 Dataset D1 R^2 on trained D1: 0.1444238661502676 .................................................. Pipeline #12: Score on D2: 0.11069141739121102 | D1-D2 diff: 2.5791968318913803 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1427763425017199 Holdout data R^2 trained on entire dataset(80%): 0.13287537059739996 Dataset D1 R^2 on trained D1: 0.1332890036752712 .................................................. Pipeline #13: Score on D2: 0.10932972532768792 | D1-D2 diff: 2.971408148653495 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10742873962872834 Holdout data R^2 trained on entire dataset(80%): 0.10125105539184243 Dataset D1 R^2 on trained D1: 0.09650196727642968 .................................................. Pipeline #14: Score on D2: 0.10705101360320379 | D1-D2 diff: 3.2302704772349045 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10742873962872834 Holdout data R^2 trained on entire dataset(80%): 0.10125105539184243 Dataset D1 R^2 on trained D1: 0.09786674773943316 .................................................. Pipeline #15: Score on D2: 0.10560342652482113 | D1-D2 diff: 3.4007249816262592 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=95), DecisionTreeRegressor(max_depth=3, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10737318325192668 Holdout data R^2 trained on entire dataset(80%): 0.10587816554449303 Dataset D1 R^2 on trained D1: 0.09812665769635487 .................................................. Pipeline #16: Score on D2: 0.10560342652482102 | D1-D2 diff: 3.4007249816262846 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=95), SelectPercentile(percentile=85), DecisionTreeRegressor(max_depth=3, min_samples_leaf=17, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1073731832519268 Holdout data R^2 trained on entire dataset(80%): 0.10587816554449303 Dataset D1 R^2 on trained D1: 0.09812665769635498 .................................................. Pipeline #17: Score on D2: 0.10459592769005222 | D1-D2 diff: 4.512141222535917 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=55), RecessiveEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11184408544744273 Holdout data R^2 trained on entire dataset(80%): 0.10122998115779125 Dataset D1 R^2 on trained D1: 0.10700843848032937 .................................................. Pipeline #18: Score on D2: 0.054584220527155725 | D1-D2 diff: 9.884385067609438 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05595587528784873 Holdout data R^2 trained on entire dataset(80%): 0.05648956604897504 Dataset D1 R^2 on trained D1: 0.05447945910721563 .................................................. Pipeline #19: Score on D2: 0.0006107562586994408 | D1-D2 diff: 23.566145702247415 Pipeline steps: VarianceThreshold(threshold=0.05), SelectPercentile(percentile=10), HeterosisEncoder(), RecessiveEncoder(), SelectPercentile(percentile=15), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.7000000000000001, min_samples_leaf=8, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0017739492075653507 Holdout data R^2 trained on entire dataset(80%): 0.001039481086830718 Dataset D1 R^2 on trained D1: 0.0006075140141700297 .................................................. ************************************************************************************** Random Seed 29 - 7 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.16417 Best score on D1-D2 diff: 4.25256 Gen 2 - Best score on D2: 0.16417 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.16417 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.17137 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.17137 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.17137 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.17141 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.17461 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.17461 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.17461 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.17461 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.17461 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.17656 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.17656 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.009691710122240904 Entire dataset(80%) R^2 trained on data (80%): 0.004605248839786968 Holdout R^2 (20%) trained on data (80%): -0.0027375877339772536 Dataset D1 R^2 on trained D1: 0.008579031093641776 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003993245227028264 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.17656217377793837 | D1-D2 diff: 1.465645100374873 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.39601241620265226 Holdout data R^2 trained on entire dataset(80%): 0.20574353781878474 Dataset D1 R^2 on trained D1: 0.3932750509165964 .................................................. Pipeline #2: Score on D2: 0.17508828843720026 | D1-D2 diff: 1.472939240605069 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=2, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.38381997881264185 Holdout data R^2 trained on entire dataset(80%): 0.21036160972960982 Dataset D1 R^2 on trained D1: 0.3875402134266541 .................................................. Pipeline #3: Score on D2: 0.17479721352799737 | D1-D2 diff: 1.5826034557825164 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34079965306781246 Holdout data R^2 trained on entire dataset(80%): 0.2063355696925374 Dataset D1 R^2 on trained D1: 0.3342057450707129 .................................................. Pipeline #4: Score on D2: 0.1726201358213988 | D1-D2 diff: 1.6011426890980616 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=4, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3305288012563987 Holdout data R^2 trained on entire dataset(80%): 0.2034166545032059 Dataset D1 R^2 on trained D1: 0.324772902319057 .................................................. Pipeline #5: Score on D2: 0.17224533176983858 | D1-D2 diff: 1.6052382360531399 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=4, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3274972619404817 Holdout data R^2 trained on entire dataset(80%): 0.20161676556721853 Dataset D1 R^2 on trained D1: 0.3228512424294028 .................................................. Pipeline #6: Score on D2: 0.16913880052001473 | D1-D2 diff: 1.635084334087785 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=7, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31431099835194853 Holdout data R^2 trained on entire dataset(80%): 0.196090208760901 Dataset D1 R^2 on trained D1: 0.3090457765706488 .................................................. Pipeline #7: Score on D2: 0.16620514272482645 | D1-D2 diff: 1.7238397831337824 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=14, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28718177219785757 Holdout data R^2 trained on entire dataset(80%): 0.18468246484159845 Dataset D1 R^2 on trained D1: 0.2794484131336167 .................................................. Pipeline #8: Score on D2: 0.16569886618726737 | D1-D2 diff: 1.7455247946830768 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=15, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2823535348839792 Holdout data R^2 trained on entire dataset(80%): 0.19154451457997212 Dataset D1 R^2 on trained D1: 0.2734187586312107 .................................................. Pipeline #9: Score on D2: 0.16470894053913465 | D1-D2 diff: 1.8354845337591443 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2644359395169602 Holdout data R^2 trained on entire dataset(80%): 0.18766837428688898 Dataset D1 R^2 on trained D1: 0.2528132359642512 .................................................. Pipeline #10: Score on D2: 0.16388101601700045 | D1-D2 diff: 1.8526733119879093 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), VarianceThreshold(threshold=0.1), HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=20, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.260670250385989 Holdout data R^2 trained on entire dataset(80%): 0.18690941823607332 Dataset D1 R^2 on trained D1: 0.24876086887869908 .................................................. Pipeline #11: Score on D2: 0.16295069283540264 | D1-D2 diff: 1.871094064139314 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.1), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25918439958723527 Holdout data R^2 trained on entire dataset(80%): 0.1888423649029174 Dataset D1 R^2 on trained D1: 0.24453704504870066 .................................................. Pipeline #12: Score on D2: 0.16222126386625613 | D1-D2 diff: 1.871890875224463 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25650431878278 Holdout data R^2 trained on entire dataset(80%): 0.1889219959463544 Dataset D1 R^2 on trained D1: 0.2436687887284671 .................................................. Pipeline #13: Score on D2: 0.16183352309871601 | D1-D2 diff: 1.876526882991094 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2538709395647788 Holdout data R^2 trained on entire dataset(80%): 0.18386711392027222 Dataset D1 R^2 on trained D1: 0.2424791529478465 .................................................. Pipeline #14: Score on D2: 0.15976924923710778 | D1-D2 diff: 1.8991362263254052 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=95), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2387562289886349 Holdout data R^2 trained on entire dataset(80%): 0.17746067076842764 Dataset D1 R^2 on trained D1: 0.23664254974487764 .................................................. Pipeline #15: Score on D2: 0.15680574031756012 | D1-D2 diff: 1.9124639483713477 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24429697224364044 Holdout data R^2 trained on entire dataset(80%): 0.17951815959076067 Dataset D1 R^2 on trained D1: 0.2315584554847312 .................................................. Pipeline #16: Score on D2: 0.154563535515164 | D1-D2 diff: 1.9791732972923963 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23204396801303673 Holdout data R^2 trained on entire dataset(80%): 0.17874790976577015 Dataset D1 R^2 on trained D1: 0.21973608469257466 .................................................. Pipeline #17: Score on D2: 0.14953611770087616 | D1-D2 diff: 1.9995480247308737 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2214680916543298 Holdout data R^2 trained on entire dataset(80%): 0.17084291983720423 Dataset D1 R^2 on trained D1: 0.21209264654295612 .................................................. Pipeline #18: Score on D2: 0.13907808306579195 | D1-D2 diff: 2.158996151501497 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=80), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20231610898967245 Holdout data R^2 trained on entire dataset(80%): 0.16821448292214747 Dataset D1 R^2 on trained D1: 0.18510294848376818 .................................................. Pipeline #19: Score on D2: 0.13256115391229473 | D1-D2 diff: 2.714112666315439 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14550555732040404 Holdout data R^2 trained on entire dataset(80%): 0.1368310374516002 Dataset D1 R^2 on trained D1: 0.15098959135499368 .................................................. Pipeline #20: Score on D2: 0.10813864710170618 | D1-D2 diff: 3.0611289779266206 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10505571185061369 Holdout data R^2 trained on entire dataset(80%): 0.0979893968925335 Dataset D1 R^2 on trained D1: 0.09674996460129992 .................................................. Pipeline #21: Score on D2: 0.10302676023444801 | D1-D2 diff: 3.557595209569014 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10505571185061369 Holdout data R^2 trained on entire dataset(80%): 0.0979893968925335 Dataset D1 R^2 on trained D1: 0.09678403998833096 .................................................. Pipeline #22: Score on D2: 0.10170558973300226 | D1-D2 diff: 4.423872770182955 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10505571185061369 Holdout data R^2 trained on entire dataset(80%): 0.0979893968925335 Dataset D1 R^2 on trained D1: 0.09909469424913553 .................................................. Pipeline #23: Score on D2: 0.045238656761903084 | D1-D2 diff: 5.070158415094708 Pipeline steps: RecessiveEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=6, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.044996171088939385 Holdout data R^2 trained on entire dataset(80%): 0.05801125873967383 Dataset D1 R^2 on trained D1: 0.04372539560556343 .................................................. Pipeline #24: Score on D2: 0.038291798546794875 | D1-D2 diff: 9.382224779220257 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03837724865308079 Holdout data R^2 trained on entire dataset(80%): 0.042286520840833175 Dataset D1 R^2 on trained D1: 0.038420854090782997 .................................................. Pipeline #25: Score on D2: 0.0007179804736084927 | D1-D2 diff: 10.177220897291688 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=10), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.001123639331240378 Holdout data R^2 trained on entire dataset(80%): -0.0006368325963013177 Dataset D1 R^2 on trained D1: 0.0006247660333670302 .................................................. Pipeline #26: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: RecessiveEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), RandomForestRegressor(max_features=0.8, min_samples_leaf=10, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 29 - 8 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.17741 Best score on D1-D2 diff: 6.50452 Gen 2 - Best score on D2: 0.18245 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.18245 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.18245 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.18245 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.18245 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.18245 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.18245 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.18245 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.18245 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.18245 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.18245 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.18245 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.18245 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.18385 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.00224891231195401 Entire dataset(80%) R^2 trained on data (80%): 0.010343916723816182 Holdout R^2 (20%) trained on data (80%): -0.0043400942202824755 Dataset D1 R^2 on trained D1: 0.011947707230547944 Combined Dataset (100%) R^2 trained on combined data (100%): 0.008296297588022084 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.18385308741343287 | D1-D2 diff: 1.510123103582004 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=5, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3746580782349077 Holdout data R^2 trained on entire dataset(80%): 0.21602986796089285 Dataset D1 R^2 on trained D1: 0.376140383373633 .................................................. Pipeline #2: Score on D2: 0.18245426217228966 | D1-D2 diff: 1.5852503163153782 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=9, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34579072853478177 Holdout data R^2 trained on entire dataset(80%): 0.20857051053854336 Dataset D1 R^2 on trained D1: 0.34080081231113357 .................................................. Pipeline #3: Score on D2: 0.18051929620454843 | D1-D2 diff: 1.6126379478338864 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=10, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33229718418423215 Holdout data R^2 trained on entire dataset(80%): 0.20520924226941517 Dataset D1 R^2 on trained D1: 0.3283799081335006 .................................................. Pipeline #4: Score on D2: 0.17638458884810404 | D1-D2 diff: 1.7443735402910299 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=14, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2936687401849817 Holdout data R^2 trained on entire dataset(80%): 0.2048574487121978 Dataset D1 R^2 on trained D1: 0.284389135512917 .................................................. Pipeline #5: Score on D2: 0.1728427781489028 | D1-D2 diff: 1.7690451340946076 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2810081609453998 Holdout data R^2 trained on entire dataset(80%): 0.20491470129585532 Dataset D1 R^2 on trained D1: 0.2749471520830856 .................................................. Pipeline #6: Score on D2: 0.1725744632996189 | D1-D2 diff: 1.8309223564004677 Pipeline steps: SelectPercentile(percentile=95), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27483021920344286 Holdout data R^2 trained on entire dataset(80%): 0.20238471388654544 Dataset D1 R^2 on trained D1: 0.261560177291824 .................................................. Pipeline #7: Score on D2: 0.1689640798727433 | D1-D2 diff: 1.8975032313953635 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=19, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2543921650085249 Holdout data R^2 trained on entire dataset(80%): 0.19652645460016915 Dataset D1 R^2 on trained D1: 0.24610235142821024 .................................................. Pipeline #8: Score on D2: 0.14969470415723807 | D1-D2 diff: 1.9009605639119322 Pipeline steps: SelectPercentile(percentile=65), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24207069233383582 Holdout data R^2 trained on entire dataset(80%): 0.19768417480199463 Dataset D1 R^2 on trained D1: 0.2262733302813491 .................................................. Pipeline #9: Score on D2: 0.14880569795662657 | D1-D2 diff: 1.905738247327638 Pipeline steps: SelectPercentile(percentile=65), VarianceThreshold(threshold=0.25), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23924295496370462 Holdout data R^2 trained on entire dataset(80%): 0.19666741712348323 Dataset D1 R^2 on trained D1: 0.22461927697514839 .................................................. Pipeline #10: Score on D2: 0.1344895870921361 | D1-D2 diff: 1.9689593360314834 Pipeline steps: SelectPercentile(percentile=95), SelectPercentile(percentile=65), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22292870088092465 Holdout data R^2 trained on entire dataset(80%): 0.18751093638967742 Dataset D1 R^2 on trained D1: 0.20102502386778187 .................................................. Pipeline #11: Score on D2: 0.12094454143476419 | D1-D2 diff: 2.0441896020100057 Pipeline steps: SelectPercentile(percentile=85), SelectPercentile(percentile=65), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2240860512144267 Holdout data R^2 trained on entire dataset(80%): 0.18795561779405368 Dataset D1 R^2 on trained D1: 0.1782129745090799 .................................................. Pipeline #12: Score on D2: 0.11798576856241683 | D1-D2 diff: 2.6705229722048833 Pipeline steps: SelectPercentile(percentile=55), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=9, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13897649557582914 Holdout data R^2 trained on entire dataset(80%): 0.14521083241449317 Dataset D1 R^2 on trained D1: 0.1376471816241035 .................................................. Pipeline #13: Score on D2: 0.10012073492265905 | D1-D2 diff: 3.24737881739227 Pipeline steps: DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09655288498304204 Holdout data R^2 trained on entire dataset(80%): 0.09557521781310263 Dataset D1 R^2 on trained D1: 0.09112848878120561 .................................................. Pipeline #14: Score on D2: 0.09113649294712522 | D1-D2 diff: 3.305829872709015 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10350703146170959 Holdout data R^2 trained on entire dataset(80%): 0.10398763944933498 Dataset D1 R^2 on trained D1: 0.09950943336593532 .................................................. Pipeline #15: Score on D2: 0.04633773458584989 | D1-D2 diff: 4.879914077853645 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=10), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.047356900340597696 Holdout data R^2 trained on entire dataset(80%): 0.05676751270264746 Dataset D1 R^2 on trained D1: 0.04810113646291958 .................................................. Pipeline #16: Score on D2: 0.043486861870106086 | D1-D2 diff: 5.274514854025093 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05131907368035682 Holdout data R^2 trained on entire dataset(80%): 0.04278993927939556 Dataset D1 R^2 on trained D1: 0.044778883907027445 .................................................. Pipeline #17: Score on D2: 0.0353179435025095 | D1-D2 diff: 5.440074991835548 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=12, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.034844848694821295 Holdout data R^2 trained on entire dataset(80%): 0.03486375287401622 Dataset D1 R^2 on trained D1: 0.03417616870142948 .................................................. Pipeline #18: Score on D2: 0.011323855878820899 | D1-D2 diff: 6.504523724995305 Pipeline steps: DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011712170185492199 Holdout data R^2 trained on entire dataset(80%): 0.006165839320719568 Dataset D1 R^2 on trained D1: 0.011882503548620194 .................................................. Pipeline #19: Score on D2: 0.005784892543315379 | D1-D2 diff: 7.482418347954337 Pipeline steps: SelectPercentile(percentile=10), HeterosisEncoder(), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.02224600735966442 Holdout data R^2 trained on entire dataset(80%): 0.03255358538902897 Dataset D1 R^2 on trained D1: 0.006103922933355377 .................................................. Pipeline #20: Score on D2: 0.0022296640714415394 | D1-D2 diff: 8.744927398515772 Pipeline steps: SelectPercentile(percentile=10), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.013761301395613046 Holdout data R^2 trained on entire dataset(80%): 0.023266645116097107 Dataset D1 R^2 on trained D1: 0.0020586723175564092 .................................................. Pipeline #21: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=10, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 29 - 9 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.19141 Best score on D1-D2 diff: 5.61942 Gen 2 - Best score on D2: 0.19351 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.19351 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.19351 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.19453 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.19453 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.19453 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.19533 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.19533 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.19635 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.19635 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.19967 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.20022 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.20214 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0009674333002962499 Entire dataset(80%) R^2 trained on data (80%): 0.009589222644272866 Holdout R^2 (20%) trained on data (80%): -0.0032734553789750542 Dataset D1 R^2 on trained D1: 0.010625724762924671 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0079064114808296 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.20214324641316894 | D1-D2 diff: 1.3994279342579166 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.46408201527665693 Holdout data R^2 trained on entire dataset(80%): 0.2284425745503189 Dataset D1 R^2 on trained D1: 0.4628773531926832 .................................................. Pipeline #2: Score on D2: 0.20188539606754075 | D1-D2 diff: 1.6132144392691699 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3543912461389812 Holdout data R^2 trained on entire dataset(80%): 0.22011979287200234 Dataset D1 R^2 on trained D1: 0.34953476591113297 .................................................. Pipeline #3: Score on D2: 0.1982721123278145 | D1-D2 diff: 1.6293973294753819 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=3, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3428837574814476 Holdout data R^2 trained on entire dataset(80%): 0.21403753021927907 Dataset D1 R^2 on trained D1: 0.34014257963897787 .................................................. Pipeline #4: Score on D2: 0.1933244533230345 | D1-D2 diff: 1.687279251425172 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31934930073503176 Holdout data R^2 trained on entire dataset(80%): 0.20313096820183785 Dataset D1 R^2 on trained D1: 0.3167065471303816 .................................................. Pipeline #5: Score on D2: 0.19196859166220392 | D1-D2 diff: 1.6917949905766143 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=13, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31574234975477133 Holdout data R^2 trained on entire dataset(80%): 0.20744503890226396 Dataset D1 R^2 on trained D1: 0.31403862447161324 .................................................. Pipeline #6: Score on D2: 0.191665149436418 | D1-D2 diff: 1.7086134997184859 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=17, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31327494426693214 Holdout data R^2 trained on entire dataset(80%): 0.20344438989891822 Dataset D1 R^2 on trained D1: 0.308999362898818 .................................................. Pipeline #7: Score on D2: 0.19025946337463318 | D1-D2 diff: 1.7466941967150866 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3024815175057596 Holdout data R^2 trained on entire dataset(80%): 0.20096975750025026 Dataset D1 R^2 on trained D1: 0.2976911739132858 .................................................. Pipeline #8: Score on D2: 0.18816276409234145 | D1-D2 diff: 1.7692946282445028 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=8, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29673670700090404 Holdout data R^2 trained on entire dataset(80%): 0.19926475549754974 Dataset D1 R^2 on trained D1: 0.29020955789098213 .................................................. Pipeline #9: Score on D2: 0.18804781612792698 | D1-D2 diff: 1.7995003221329513 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=20, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28792428723569774 Holdout data R^2 trained on entire dataset(80%): 0.1980495843870027 Dataset D1 R^2 on trained D1: 0.2834135346063581 .................................................. Pipeline #10: Score on D2: 0.18636494884576527 | D1-D2 diff: 1.803611396380548 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=20, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2895701544582323 Holdout data R^2 trained on entire dataset(80%): 0.20167128217705432 Dataset D1 R^2 on trained D1: 0.28086414555783545 .................................................. Pipeline #11: Score on D2: 0.18525183120102107 | D1-D2 diff: 1.8072985394199987 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=18, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28462495058007564 Holdout data R^2 trained on entire dataset(80%): 0.19654691775124278 Dataset D1 R^2 on trained D1: 0.2789822180640036 .................................................. Pipeline #12: Score on D2: 0.1839593921570314 | D1-D2 diff: 1.811666621705385 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=18, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28548787754926075 Holdout data R^2 trained on entire dataset(80%): 0.19147525459338632 Dataset D1 R^2 on trained D1: 0.2767890753492588 .................................................. Pipeline #13: Score on D2: 0.18331483496229373 | D1-D2 diff: 1.825911579457422 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28264343582260554 Holdout data R^2 trained on entire dataset(80%): 0.19504693717097366 Dataset D1 R^2 on trained D1: 0.2732813771099286 .................................................. Pipeline #14: Score on D2: 0.18282804908952266 | D1-D2 diff: 1.882573353318146 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=20, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27411948832239885 Holdout data R^2 trained on entire dataset(80%): 0.1934512329564435 Dataset D1 R^2 on trained D1: 0.2624425855976985 .................................................. Pipeline #15: Score on D2: 0.18267718031149904 | D1-D2 diff: 1.933490169490067 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26558599567618346 Holdout data R^2 trained on entire dataset(80%): 0.19116651988705335 Dataset D1 R^2 on trained D1: 0.25423088540976857 .................................................. Pipeline #16: Score on D2: 0.17668251131533552 | D1-D2 diff: 1.9840362556751079 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25138278013220583 Holdout data R^2 trained on entire dataset(80%): 0.18512419656779255 Dataset D1 R^2 on trained D1: 0.24121844293595884 .................................................. Pipeline #17: Score on D2: 0.17641965431087137 | D1-D2 diff: 1.987650524991961 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=18, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24940579061964052 Holdout data R^2 trained on entire dataset(80%): 0.18474997138979687 Dataset D1 R^2 on trained D1: 0.24048746579630575 .................................................. Pipeline #18: Score on D2: 0.17360486585839474 | D1-D2 diff: 2.03709797439537 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2408103298102231 Holdout data R^2 trained on entire dataset(80%): 0.18443383221235787 Dataset D1 R^2 on trained D1: 0.23167493354846302 .................................................. Pipeline #19: Score on D2: 0.1726641821269912 | D1-D2 diff: 2.0406933051084453 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24195134809794538 Holdout data R^2 trained on entire dataset(80%): 0.18518823365021264 Dataset D1 R^2 on trained D1: 0.23032609442476515 .................................................. Pipeline #20: Score on D2: 0.1675091613969938 | D1-D2 diff: 2.0613877041977973 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2349151586700432 Holdout data R^2 trained on entire dataset(80%): 0.1813317693515193 Dataset D1 R^2 on trained D1: 0.22289022281788606 .................................................. Pipeline #21: Score on D2: 0.1633340596145203 | D1-D2 diff: 2.0952375338676563 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22836099350344985 Holdout data R^2 trained on entire dataset(80%): 0.1796347406270984 Dataset D1 R^2 on trained D1: 0.21522206046263836 .................................................. Pipeline #22: Score on D2: 0.14321519407127226 | D1-D2 diff: 2.0954841919709266 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20908398417041085 Holdout data R^2 trained on entire dataset(80%): 0.16407975148672715 Dataset D1 R^2 on trained D1: 0.19507876841915472 .................................................. Pipeline #23: Score on D2: 0.14266262779781602 | D1-D2 diff: 2.1284110283646998 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2029918002467921 Holdout data R^2 trained on entire dataset(80%): 0.1539704682344739 Dataset D1 R^2 on trained D1: 0.1913905617435675 .................................................. Pipeline #24: Score on D2: 0.13751111176981057 | D1-D2 diff: 2.1651838965591814 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19764047883352232 Holdout data R^2 trained on entire dataset(80%): 0.15323405607079743 Dataset D1 R^2 on trained D1: 0.18301210181237404 .................................................. Pipeline #25: Score on D2: 0.1327016213062957 | D1-D2 diff: 2.329932657361749 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18526622825945782 Holdout data R^2 trained on entire dataset(80%): 0.15245916090927314 Dataset D1 R^2 on trained D1: 0.16663495479482382 .................................................. Pipeline #26: Score on D2: 0.12220266564986171 | D1-D2 diff: 2.413473691159666 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19004006216940006 Holdout data R^2 trained on entire dataset(80%): 0.15557569888910738 Dataset D1 R^2 on trained D1: 0.15167603075009028 .................................................. Pipeline #27: Score on D2: 0.12161711041810541 | D1-D2 diff: 2.4819251711626116 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18329908201847966 Holdout data R^2 trained on entire dataset(80%): 0.1528539720874038 Dataset D1 R^2 on trained D1: 0.1479710329892967 .................................................. Pipeline #28: Score on D2: 0.1080814264158515 | D1-D2 diff: 4.461825876485439 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10970821412856224 Holdout data R^2 trained on entire dataset(80%): 0.08770516857229549 Dataset D1 R^2 on trained D1: 0.11060461393953991 .................................................. Pipeline #29: Score on D2: 0.09201643556508465 | D1-D2 diff: 4.946211227028195 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=20), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0915048303404068 Holdout data R^2 trained on entire dataset(80%): 0.08087251903383319 Dataset D1 R^2 on trained D1: 0.09034569366440237 .................................................. Pipeline #30: Score on D2: 0.08760833532978529 | D1-D2 diff: 5.024735342388045 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=15), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08970636665102294 Holdout data R^2 trained on entire dataset(80%): 0.07040224957142738 Dataset D1 R^2 on trained D1: 0.08917706182628626 .................................................. Pipeline #31: Score on D2: 0.08026649708302169 | D1-D2 diff: 5.067226308164917 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=20), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08188095744575241 Holdout data R^2 trained on entire dataset(80%): 0.06679912144037448 Dataset D1 R^2 on trained D1: 0.07874973034321331 .................................................. Pipeline #32: Score on D2: 0.07790824762284476 | D1-D2 diff: 6.213300038224791 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=20), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07787456630658207 Holdout data R^2 trained on entire dataset(80%): 0.06698174574296756 Dataset D1 R^2 on trained D1: 0.07723726589170377 .................................................. Pipeline #33: Score on D2: 0.06154229609274986 | D1-D2 diff: 7.506133769478754 Pipeline steps: SelectPercentile(percentile=70), UnderDominanceEncoder(), SelectPercentile(percentile=60), DecisionTreeRegressor(max_depth=2, min_samples_leaf=16, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0635408360656512 Holdout data R^2 trained on entire dataset(80%): 0.056641170132596286 Dataset D1 R^2 on trained D1: 0.061227278505736504 .................................................. Pipeline #34: Score on D2: 0.035793130866718315 | D1-D2 diff: 8.344824079252014 Pipeline steps: RecessiveEncoder(), HeterosisEncoder(), SelectPercentile(percentile=15), RandomForestRegressor(max_features=0.55, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03648398934561592 Holdout data R^2 trained on entire dataset(80%): 0.017396352807158788 Dataset D1 R^2 on trained D1: 0.03599935109236718 .................................................. Pipeline #35: Score on D2: 0.0022296640647814225 | D1-D2 diff: 8.744927411333736 Pipeline steps: SelectPercentile(percentile=10), UnderDominanceEncoder(), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002056070482482353 Holdout data R^2 trained on entire dataset(80%): 0.005000250533414019 Dataset D1 R^2 on trained D1: 0.0020586723118988237 .................................................. Pipeline #36: Score on D2: 0.0002795936709660829 | D1-D2 diff: 12.551648112195261 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=20), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00603070366128422 Holdout data R^2 trained on entire dataset(80%): -0.006590660621529132 Dataset D1 R^2 on trained D1: 0.0003198836442221964 .................................................. Pipeline #37: Score on D2: -3.2015594250811574e-05 | D1-D2 diff: 13.480151627526027 Pipeline steps: DominantEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021467281316e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516763899763 Dataset D1 R^2 on trained D1: -1.73106056000627e-06 .................................................. ************************************************************************************** Random Seed 30 - 0 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00307 Best score on D1-D2 diff: 3.52237 Gen 2 - Best score on D2: -0.00005 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: -0.00005 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: -0.00003 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.00016 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.019782885776708836 Entire dataset(80%) R^2 trained on data (80%): 0.0042146812855679006 Holdout R^2 (20%) trained on data (80%): -0.012930726470369747 Dataset D1 R^2 on trained D1: 0.012834010629451775 Combined Dataset (100%) R^2 trained on combined data (100%): 0.002179279432735215 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.00015961647802165135 | D1-D2 diff: 2.88770574702274 Pipeline steps: RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.013656569820173492 Holdout data R^2 trained on entire dataset(80%): -0.0034873022094468187 Dataset D1 R^2 on trained D1: 0.014540588827979373 .................................................. Pipeline #2: Score on D2: -3.2015594250811574e-05 | D1-D2 diff: 13.480151627526027 Pipeline steps: RecessiveEncoder(), DominantEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=12, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021467281316e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516763899763 Dataset D1 R^2 on trained D1: -1.73106056000627e-06 .................................................. ************************************************************************************** Random Seed 30 - 1 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.04654 Best score on D1-D2 diff: 3.52237 Gen 2 - Best score on D2: 0.04655 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.05779 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.05779 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.05779 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.05779 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.05978 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.016440883177751164 Entire dataset(80%) R^2 trained on data (80%): 0.004064861217282112 Holdout R^2 (20%) trained on data (80%): -0.011953682433126334 Dataset D1 R^2 on trained D1: 0.01224600960287514 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0022266751046370326 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.05978332038276668 | D1-D2 diff: 1.7666917429718347 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.8, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.181526730221383 Holdout data R^2 trained on entire dataset(80%): 0.06596029225243816 Dataset D1 R^2 on trained D1: 0.16243283093028416 .................................................. Pipeline #2: Score on D2: 0.05842723977646225 | D1-D2 diff: 1.767960990641902 Pipeline steps: VarianceThreshold(threshold=0.05), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), RandomForestRegressor(max_features=0.8, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1754197508416273 Holdout data R^2 trained on entire dataset(80%): 0.07013307230648236 Dataset D1 R^2 on trained D1: 0.16078229265999788 .................................................. Pipeline #3: Score on D2: 0.056853214745364644 | D1-D2 diff: 1.7999766069509746 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=20, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.171024439067624 Holdout data R^2 trained on entire dataset(80%): 0.06330296696797055 Dataset D1 R^2 on trained D1: 0.1521180358704246 .................................................. Pipeline #4: Score on D2: 0.05577075869428072 | D1-D2 diff: 1.8230515206544717 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), VarianceThreshold(threshold=0.1), RandomForestRegressor(max_features=0.55, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1637864087593407 Holdout data R^2 trained on entire dataset(80%): 0.06308181871638674 Dataset D1 R^2 on trained D1: 0.1463031998177453 .................................................. Pipeline #5: Score on D2: 0.05424861941484982 | D1-D2 diff: 1.8648082958976426 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1497971601730469 Holdout data R^2 trained on entire dataset(80%): 0.052917016878859635 Dataset D1 R^2 on trained D1: 0.13694056873978278 .................................................. Pipeline #6: Score on D2: 0.05327218287564428 | D1-D2 diff: 1.8842004949172115 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14223091728306925 Holdout data R^2 trained on entire dataset(80%): 0.05253740337934221 Dataset D1 R^2 on trained D1: 0.1326120640798134 .................................................. Pipeline #7: Score on D2: 0.05093992792178215 | D1-D2 diff: 1.9145840799753806 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13881046319178358 Holdout data R^2 trained on entire dataset(80%): 0.05258063357812215 Dataset D1 R^2 on trained D1: 0.12536208035374352 .................................................. Pipeline #8: Score on D2: 0.04584631239690351 | D1-D2 diff: 1.9311831220977156 Pipeline steps: VarianceThreshold(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12953388548431255 Holdout data R^2 trained on entire dataset(80%): 0.049956765116777935 Dataset D1 R^2 on trained D1: 0.11774255121473887 .................................................. Pipeline #9: Score on D2: 0.04369368241825511 | D1-D2 diff: 1.9412257471666863 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1353847127039084 Holdout data R^2 trained on entire dataset(80%): 0.05671815650547729 Dataset D1 R^2 on trained D1: 0.11411365124239403 .................................................. Pipeline #10: Score on D2: 0.039852989847123044 | D1-D2 diff: 1.9612593504238358 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=19, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12399293898786834 Holdout data R^2 trained on entire dataset(80%): 0.0495977996466761 Dataset D1 R^2 on trained D1: 0.10743947967321821 .................................................. Pipeline #11: Score on D2: 0.03951056623300109 | D1-D2 diff: 1.9647944087617761 Pipeline steps: VarianceThreshold(threshold=0.25), VarianceThreshold(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12581394826590808 Holdout data R^2 trained on entire dataset(80%): 0.05683047328564683 Dataset D1 R^2 on trained D1: 0.10661196071646528 .................................................. Pipeline #12: Score on D2: 0.03524894239714382 | D1-D2 diff: 2.0841522211270185 Pipeline steps: VarianceThreshold(), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10109199502823074 Holdout data R^2 trained on entire dataset(80%): 0.038530368304839246 Dataset D1 R^2 on trained D1: 0.08824972200459535 .................................................. Pipeline #13: Score on D2: 0.025280358002865877 | D1-D2 diff: 2.1016370115955305 Pipeline steps: VarianceThreshold(), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.1, min_samples_leaf=20, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08347901950594128 Holdout data R^2 trained on entire dataset(80%): 0.024782419433128866 Dataset D1 R^2 on trained D1: 0.07653924446858751 .................................................. Pipeline #14: Score on D2: 0.023014119223021323 | D1-D2 diff: 2.1219237150097974 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=13, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08286012980125324 Holdout data R^2 trained on entire dataset(80%): 0.021879584486832138 Dataset D1 R^2 on trained D1: 0.07234069109652874 .................................................. Pipeline #15: Score on D2: 0.022110974108635495 | D1-D2 diff: 2.1336988990124777 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08564024795788194 Holdout data R^2 trained on entire dataset(80%): 0.03278563936050316 Dataset D1 R^2 on trained D1: 0.07035765787850001 .................................................. Pipeline #16: Score on D2: 0.020584784515020438 | D1-D2 diff: 2.1494763410937923 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=14, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08170742819108845 Holdout data R^2 trained on entire dataset(80%): 0.02498846748482486 Dataset D1 R^2 on trained D1: 0.06743044029400391 .................................................. Pipeline #17: Score on D2: 0.020538821169377663 | D1-D2 diff: 2.3448455179518475 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06514172252134076 Holdout data R^2 trained on entire dataset(80%): 0.023834131643432066 Dataset D1 R^2 on trained D1: 0.053617111597401235 .................................................. Pipeline #18: Score on D2: 0.018773159481795032 | D1-D2 diff: 2.3572314262718983 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.059569784617724886 Holdout data R^2 trained on entire dataset(80%): 0.02097654166515539 Dataset D1 R^2 on trained D1: 0.05116168000806709 .................................................. Pipeline #19: Score on D2: 0.017674298066475136 | D1-D2 diff: 2.373646658567783 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.055613421961543064 Holdout data R^2 trained on entire dataset(80%): 0.019384000361771014 Dataset D1 R^2 on trained D1: 0.0491761233788135 .................................................. Pipeline #20: Score on D2: 0.0010657949894351537 | D1-D2 diff: 2.463614084486211 Pipeline steps: OverDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=14, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.026234784401425237 Holdout data R^2 trained on entire dataset(80%): -0.0012562170135737016 Dataset D1 R^2 on trained D1: 0.02821201014163044 .................................................. Pipeline #21: Score on D2: 0.0004724847366135432 | D1-D2 diff: 2.715394299549656 Pipeline steps: DominantEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.019045745186736074 Holdout data R^2 trained on entire dataset(80%): -0.002969102908579657 Dataset D1 R^2 on trained D1: 0.018866154806222668 .................................................. Pipeline #22: Score on D2: 0.00020334091231721807 | D1-D2 diff: 2.719179111850132 Pipeline steps: DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.018006929332390786 Holdout data R^2 trained on entire dataset(80%): -0.0018268326829520287 Dataset D1 R^2 on trained D1: 0.018494816353021748 .................................................. Pipeline #23: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=3, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 30 - 2 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.08137 Best score on D1-D2 diff: 3.52237 Gen 2 - Best score on D2: 0.09758 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.10322 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.10549 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.10549 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.10549 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.10549 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.10549 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.10549 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.10792 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.10792 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.10792 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.10792 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.10792 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.10792 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.10935 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.10935 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.10935 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.10935 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.018323370431920782 Entire dataset(80%) R^2 trained on data (80%): 0.0040022422540453295 Holdout R^2 (20%) trained on data (80%): -0.01179091476921057 Dataset D1 R^2 on trained D1: 0.012974586466668048 Combined Dataset (100%) R^2 trained on combined data (100%): 0.002100994268662726 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.10934954124224006 | D1-D2 diff: 1.7762043322430026 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), HeterosisEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.8, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22083278633002057 Holdout data R^2 trained on entire dataset(80%): 0.12916102426133658 Dataset D1 R^2 on trained D1: 0.2098176668859142 .................................................. Pipeline #2: Score on D2: 0.1039826843056767 | D1-D2 diff: 1.839007538987831 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.1), VarianceThreshold(threshold=0.2), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2060248016132149 Holdout data R^2 trained on entire dataset(80%): 0.12525234810718522 Dataset D1 R^2 on trained D1: 0.19141378814422094 .................................................. Pipeline #3: Score on D2: 0.10291786632458522 | D1-D2 diff: 1.8464519755840947 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20464637516935857 Holdout data R^2 trained on entire dataset(80%): 0.12363520937323946 Dataset D1 R^2 on trained D1: 0.18894747229558873 .................................................. Pipeline #4: Score on D2: 0.0976662809541543 | D1-D2 diff: 1.864950574705649 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), FeatureEncodingFrequencySelector(threshold=0.1), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), RandomForestRegressor(max_features=0.4, min_samples_leaf=16, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19783127909708687 Holdout data R^2 trained on entire dataset(80%): 0.12275165756717776 Dataset D1 R^2 on trained D1: 0.1803329985848432 .................................................. Pipeline #5: Score on D2: 0.09335999009053109 | D1-D2 diff: 1.9013474750049715 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), FeatureEncodingFrequencySelector(threshold=0.1), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), RandomForestRegressor(max_features=0.3, min_samples_leaf=16, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18538318517553465 Holdout data R^2 trained on entire dataset(80%): 0.11809311160559888 Dataset D1 R^2 on trained D1: 0.1698763023502462 .................................................. Pipeline #6: Score on D2: 0.07557497137147451 | D1-D2 diff: 1.9113704164016037 Pipeline steps: VarianceThreshold(threshold=0.25), FeatureEncodingFrequencySelector(threshold=0.3), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17188461869477545 Holdout data R^2 trained on entire dataset(80%): 0.09820236997166998 Dataset D1 R^2 on trained D1: 0.1504989032977173 .................................................. Pipeline #7: Score on D2: 0.07552733395703215 | D1-D2 diff: 2.064558706013791 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.1), VarianceThreshold(threshold=0.2), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.2, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1411901218639514 Holdout data R^2 trained on entire dataset(80%): 0.08488734651564478 Dataset D1 R^2 on trained D1: 0.13056893442284512 .................................................. Pipeline #8: Score on D2: 0.07097415953788766 | D1-D2 diff: 2.0816620274980644 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.2, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13865014196651826 Holdout data R^2 trained on entire dataset(80%): 0.08544495440273836 Dataset D1 R^2 on trained D1: 0.124229003862801 .................................................. Pipeline #9: Score on D2: 0.06087253985205077 | D1-D2 diff: 2.655779453422781 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), VarianceThreshold(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=7, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11718329313397313 Holdout data R^2 trained on entire dataset(80%): 0.09949685610766268 Dataset D1 R^2 on trained D1: 0.08097420218498041 .................................................. Pipeline #10: Score on D2: 0.025888370755900558 | D1-D2 diff: 2.912442228970006 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06745977368279588 Holdout data R^2 trained on entire dataset(80%): 0.04844849927829309 Dataset D1 R^2 on trained D1: 0.03978696006439608 .................................................. Pipeline #11: Score on D2: 0.0006314031218213056 | D1-D2 diff: 4.275432795470044 Pipeline steps: VarianceThreshold(threshold=0.2), SelectPercentile(percentile=80), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004396751902964069 Holdout data R^2 trained on entire dataset(80%): -0.005487901407287277 Dataset D1 R^2 on trained D1: 0.003624216682369008 .................................................. Pipeline #12: Score on D2: 4.65399183920967e-05 | D1-D2 diff: 4.2877050294249655 Pipeline steps: VarianceThreshold(threshold=0.3), SelectPercentile(percentile=80), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0037977014830017364 Holdout data R^2 trained on entire dataset(80%): -0.0034066183340151213 Dataset D1 R^2 on trained D1: 0.003005236278649459 .................................................. Pipeline #13: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: RecessiveEncoder(), DominantEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=12, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 30 - 3 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.12638 Best score on D1-D2 diff: 5.57012 Gen 2 - Best score on D2: 0.13202 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.13202 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.13643 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.13656 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.13849 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.13849 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.13849 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.13849 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.13981 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.13981 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.13981 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.13981 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.13981 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.14501 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.14501 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.14501 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.14501 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.013747915141240519 Entire dataset(80%) R^2 trained on data (80%): 0.004685449148888465 Holdout R^2 (20%) trained on data (80%): -0.009203978683655523 Dataset D1 R^2 on trained D1: 0.012367063126967892 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003202399827875957 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.14500780660667112 | D1-D2 diff: 1.618055822845624 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), VarianceThreshold(threshold=0.05), RandomForestRegressor(max_features=0.55, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29670574494789415 Holdout data R^2 trained on entire dataset(80%): 0.14928376813603472 Dataset D1 R^2 on trained D1: 0.29089796550591573 .................................................. Pipeline #2: Score on D2: 0.13880201619215782 | D1-D2 diff: 1.72524797963279 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2606328743990022 Holdout data R^2 trained on entire dataset(80%): 0.14579436983164917 Dataset D1 R^2 on trained D1: 0.25167600951012703 .................................................. Pipeline #3: Score on D2: 0.13848615781057672 | D1-D2 diff: 1.756555879687307 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=14, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25254655641380575 Holdout data R^2 trained on entire dataset(80%): 0.1448873612995043 Dataset D1 R^2 on trained D1: 0.24352553048927827 .................................................. Pipeline #4: Score on D2: 0.13831626316622292 | D1-D2 diff: 1.8519619937526839 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=18, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22912098083810983 Holdout data R^2 trained on entire dataset(80%): 0.14282109470278082 Dataset D1 R^2 on trained D1: 0.22332659685037937 .................................................. Pipeline #5: Score on D2: 0.13788083680126983 | D1-D2 diff: 1.852566843259797 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), OverDominanceEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=18, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2334171502231821 Holdout data R^2 trained on entire dataset(80%): 0.14503956869193124 Dataset D1 R^2 on trained D1: 0.22278020383940034 .................................................. Pipeline #6: Score on D2: 0.13782547835255377 | D1-D2 diff: 1.938297132265925 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), VarianceThreshold(threshold=0.05), RandomForestRegressor(max_features=0.55, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2173815697801097 Holdout data R^2 trained on entire dataset(80%): 0.1435568513768799 Dataset D1 R^2 on trained D1: 0.20867200890514892 .................................................. Pipeline #7: Score on D2: 0.12542954996938172 | D1-D2 diff: 2.008317492790653 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), VarianceThreshold(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.4, min_samples_leaf=20, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20039093470314318 Holdout data R^2 trained on entire dataset(80%): 0.1381540459030124 Dataset D1 R^2 on trained D1: 0.18690058359500195 .................................................. Pipeline #8: Score on D2: 0.10198549430152692 | D1-D2 diff: 2.4059990714024115 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08142829606745028 Holdout data R^2 trained on entire dataset(80%): 0.08741069378595323 Dataset D1 R^2 on trained D1: 0.1318268245160189 .................................................. Pipeline #9: Score on D2: 0.10193182022909153 | D1-D2 diff: 2.409366786212123 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08064228215733982 Holdout data R^2 trained on entire dataset(80%): 0.08550261002211546 Dataset D1 R^2 on trained D1: 0.13160665594313103 .................................................. Pipeline #10: Score on D2: 0.09461459979485931 | D1-D2 diff: 3.377290804221797 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=6, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004284799440348097 Holdout data R^2 trained on entire dataset(80%): -0.007951709013582553 Dataset D1 R^2 on trained D1: 0.10230105622861319 .................................................. Pipeline #11: Score on D2: 0.004661511157400766 | D1-D2 diff: 3.698397788236426 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), SelectPercentile(percentile=45), DecisionTreeRegressor(max_depth=3, min_samples_leaf=10, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0067991274495364795 Holdout data R^2 trained on entire dataset(80%): -0.013368639593379639 Dataset D1 R^2 on trained D1: 0.01000648417520611 .................................................. Pipeline #12: Score on D2: 0.0037191089740840066 | D1-D2 diff: 5.080659331163765 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(), SelectPercentile(percentile=55), VarianceThreshold(threshold=0.35), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004814062355548487 Holdout data R^2 trained on entire dataset(80%): -0.004562591168073338 Dataset D1 R^2 on trained D1: 0.005219898181332505 .................................................. Pipeline #13: Score on D2: 0.0006283195916512163 | D1-D2 diff: 5.570123288910373 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.001597262540473765 Holdout data R^2 trained on entire dataset(80%): -0.0027326882841101074 Dataset D1 R^2 on trained D1: 0.0016671407060799837 .................................................. Pipeline #14: Score on D2: 0.0005319411195822132 | D1-D2 diff: 7.357453941014449 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=5), VarianceThreshold(threshold=0.35), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0007347747650490177 Holdout data R^2 trained on entire dataset(80%): -0.00011914963059433958 Dataset D1 R^2 on trained D1: 0.000873204581342768 .................................................. Pipeline #15: Score on D2: 0.00025940572415850127 | D1-D2 diff: 11.871249196740377 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=5), HeterosisEncoder(), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00025894648255930797 Holdout data R^2 trained on entire dataset(80%): -0.0001625149838462292 Dataset D1 R^2 on trained D1: 0.00020905400381276973 .................................................. Pipeline #16: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 30 - 4 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.13817 Best score on D1-D2 diff: 5.09866 Gen 2 - Best score on D2: 0.14404 Best score on D1-D2 diff: 5.09866 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.013994626549778344 Entire dataset(80%) R^2 trained on data (80%): 0.00463591910297112 Holdout R^2 (20%) trained on data (80%): -0.00904423884437544 Dataset D1 R^2 on trained D1: 0.012057359183685223 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0033005752288244317 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.14404358738771927 | D1-D2 diff: 1.4556171712359607 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=6, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.36637921843733845 Holdout data R^2 trained on entire dataset(80%): 0.16234985675649716 Dataset D1 R^2 on trained D1: 0.3667903083531031 .................................................. Pipeline #2: Score on D2: 0.13944055902722197 | D1-D2 diff: 1.6884816621199483 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2672391302542835 Holdout data R^2 trained on entire dataset(80%): 0.15715210752404019 Dataset D1 R^2 on trained D1: 0.2624715739792345 .................................................. Pipeline #3: Score on D2: 0.08322496403108537 | D1-D2 diff: 2.4770383277166785 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0972339534566673 Holdout data R^2 trained on entire dataset(80%): 0.11959356637623941 Dataset D1 R^2 on trained D1: 0.10978747298206759 .................................................. Pipeline #4: Score on D2: 0.07585325761687978 | D1-D2 diff: 2.61427072675369 Pipeline steps: DecisionTreeRegressor(max_depth=3, min_samples_leaf=18, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08913006344804542 Holdout data R^2 trained on entire dataset(80%): 0.09786360567646557 Dataset D1 R^2 on trained D1: 0.09726232555326864 .................................................. Pipeline #5: Score on D2: 0.03401999639061137 | D1-D2 diff: 3.4742719192259464 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.037694953292289224 Holdout data R^2 trained on entire dataset(80%): 0.025858973110866823 Dataset D1 R^2 on trained D1: 0.04088348275699327 .................................................. Pipeline #6: Score on D2: 0.0014282325915688787 | D1-D2 diff: 5.098664324292568 Pipeline steps: DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0022488334179161518 Holdout data R^2 trained on entire dataset(80%): 0.001170296371741375 Dataset D1 R^2 on trained D1: 0.0029079347837890346 .................................................. ************************************************************************************** Random Seed 30 - 5 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.15038 Best score on D1-D2 diff: 4.16228 Gen 2 - Best score on D2: 0.15074 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.15313 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15313 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.15313 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.15362 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15644 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.15644 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.15644 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.15644 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.15676 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.15676 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.16170 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.017247487903857994 Entire dataset(80%) R^2 trained on data (80%): 0.004516632142700372 Holdout R^2 (20%) trained on data (80%): -0.009287324176875167 Dataset D1 R^2 on trained D1: 0.013630827839830162 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003174676336547222 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.16170414367339636 | D1-D2 diff: 1.5066709383581398 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=3, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.36879732052126923 Holdout data R^2 trained on entire dataset(80%): 0.18793383791161422 Dataset D1 R^2 on trained D1: 0.3557598216022875 .................................................. Pipeline #2: Score on D2: 0.15687321391132125 | D1-D2 diff: 1.550157341617588 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=3, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33150410364228533 Holdout data R^2 trained on entire dataset(80%): 0.17801267393730602 Dataset D1 R^2 on trained D1: 0.330052870489006 .................................................. Pipeline #3: Score on D2: 0.151084103253752 | D1-D2 diff: 1.6072300283498697 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), VarianceThreshold(threshold=0.05), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=11, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3095828532826074 Holdout data R^2 trained on entire dataset(80%): 0.18102399920404788 Dataset D1 R^2 on trained D1: 0.3009448348871424 .................................................. Pipeline #4: Score on D2: 0.15105258808012345 | D1-D2 diff: 1.6410520592248523 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=12, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29727449132711714 Holdout data R^2 trained on entire dataset(80%): 0.18090453978756849 Dataset D1 R^2 on trained D1: 0.28893553799514504 .................................................. Pipeline #5: Score on D2: 0.15060144551655708 | D1-D2 diff: 1.7459958951338637 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26690644288860854 Holdout data R^2 trained on entire dataset(80%): 0.18070417430926955 Dataset D1 R^2 on trained D1: 0.2582051261065841 .................................................. Pipeline #6: Score on D2: 0.14810974361526086 | D1-D2 diff: 1.7625304303338802 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2621589895167633 Holdout data R^2 trained on entire dataset(80%): 0.1803931980746304 Dataset D1 R^2 on trained D1: 0.25173210966238013 .................................................. Pipeline #7: Score on D2: 0.14699242868655282 | D1-D2 diff: 1.7796850933725248 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=20, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25743395379972445 Holdout data R^2 trained on entire dataset(80%): 0.17853039521606118 Dataset D1 R^2 on trained D1: 0.2466768630174917 .................................................. Pipeline #8: Score on D2: 0.14649674685748726 | D1-D2 diff: 1.783509901361066 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2556790488213393 Holdout data R^2 trained on entire dataset(80%): 0.1759022017571935 Dataset D1 R^2 on trained D1: 0.2453288190284001 .................................................. Pipeline #9: Score on D2: 0.1451644859977015 | D1-D2 diff: 1.8441868325113424 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2382597295686163 Holdout data R^2 trained on entire dataset(80%): 0.1689040227921278 Dataset D1 R^2 on trained D1: 0.23161753862251022 .................................................. Pipeline #10: Score on D2: 0.1427630943667224 | D1-D2 diff: 1.863883929731361 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=19, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23621774441199506 Holdout data R^2 trained on entire dataset(80%): 0.1692803337066041 Dataset D1 R^2 on trained D1: 0.2256192052469289 .................................................. Pipeline #11: Score on D2: 0.14051467109848903 | D1-D2 diff: 1.880465888720562 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23193526160169264 Holdout data R^2 trained on entire dataset(80%): 0.17234627151703696 Dataset D1 R^2 on trained D1: 0.22048670855823072 .................................................. Pipeline #12: Score on D2: 0.14043048760492804 | D1-D2 diff: 1.8979715881992694 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22511994354474962 Holdout data R^2 trained on entire dataset(80%): 0.16933529922164614 Dataset D1 R^2 on trained D1: 0.21749264661189627 .................................................. Pipeline #13: Score on D2: 0.13789998267624193 | D1-D2 diff: 1.9083243618069323 Pipeline steps: HeterosisEncoder(), VarianceThreshold(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22388762377808424 Holdout data R^2 trained on entire dataset(80%): 0.1676592043812144 Dataset D1 R^2 on trained D1: 0.21330343350179104 .................................................. Pipeline #14: Score on D2: 0.1367362264066525 | D1-D2 diff: 1.9381631767343503 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21755016590532095 Holdout data R^2 trained on entire dataset(80%): 0.16920291507637375 Dataset D1 R^2 on trained D1: 0.2076023451315887 .................................................. Pipeline #15: Score on D2: 0.1354951354795042 | D1-D2 diff: 1.941863462682175 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22055636742730222 Holdout data R^2 trained on entire dataset(80%): 0.17000758497798896 Dataset D1 R^2 on trained D1: 0.20582264509520143 .................................................. Pipeline #16: Score on D2: 0.13481827810557023 | D1-D2 diff: 1.9827843372152512 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21189697450972367 Holdout data R^2 trained on entire dataset(80%): 0.1632304194976385 Dataset D1 R^2 on trained D1: 0.19951735460049003 .................................................. Pipeline #17: Score on D2: 0.13238523675966796 | D1-D2 diff: 2.0049602558876454 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2128995012493321 Holdout data R^2 trained on entire dataset(80%): 0.16839923695790426 Dataset D1 R^2 on trained D1: 0.19426903018359465 .................................................. Pipeline #18: Score on D2: 0.10219708227700797 | D1-D2 diff: 2.015062094625043 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.1, min_samples_leaf=11, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18317256875860355 Holdout data R^2 trained on entire dataset(80%): 0.1315716740090881 Dataset D1 R^2 on trained D1: 0.16284924140281787 .................................................. Pipeline #19: Score on D2: 0.10164909900749408 | D1-D2 diff: 2.201903487874281 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09447265524807935 Holdout data R^2 trained on entire dataset(80%): 0.09417598323394061 Dataset D1 R^2 on trained D1: 0.1441900194581185 .................................................. Pipeline #20: Score on D2: 0.09436624928014115 | D1-D2 diff: 2.2276559956352306 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=90), RandomForestRegressor(max_features=0.1, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15745227699084496 Holdout data R^2 trained on entire dataset(80%): 0.12293940020926852 Dataset D1 R^2 on trained D1: 0.13497386549432988 .................................................. Pipeline #21: Score on D2: 0.08264798100042581 | D1-D2 diff: 2.460386945278362 Pipeline steps: SelectPercentile(percentile=50), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08167833301701877 Holdout data R^2 trained on entire dataset(80%): 0.1226441360294458 Dataset D1 R^2 on trained D1: 0.10993690073756568 .................................................. Pipeline #22: Score on D2: 0.07236622903347922 | D1-D2 diff: 2.7512071139112404 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08167833301701899 Holdout data R^2 trained on entire dataset(80%): 0.1226441360294458 Dataset D1 R^2 on trained D1: 0.08982070669868958 .................................................. Pipeline #23: Score on D2: 0.043337291786228915 | D1-D2 diff: 2.863893218167374 Pipeline steps: VarianceThreshold(threshold=0.15), DecisionTreeRegressor(max_depth=2, min_samples_leaf=9, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.051452195495765074 Holdout data R^2 trained on entire dataset(80%): 0.05668066872410382 Dataset D1 R^2 on trained D1: 0.058202558846625396 .................................................. Pipeline #24: Score on D2: 0.014731462454565358 | D1-D2 diff: 4.162276897315696 Pipeline steps: SelectPercentile(percentile=90), RecessiveEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=18, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01649029065836316 Holdout data R^2 trained on entire dataset(80%): 0.00786731077951286 Dataset D1 R^2 on trained D1: 0.01806324099110579 .................................................. Pipeline #25: Score on D2: 0.0019622207799663283 | D1-D2 diff: 4.353394428705209 Pipeline steps: SelectPercentile(percentile=35), HeterosisEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), RecessiveEncoder(), SelectPercentile(percentile=95), SelectPercentile(percentile=45), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03584605583513101 Holdout data R^2 trained on entire dataset(80%): 0.04392304700866145 Dataset D1 R^2 on trained D1: 0.004746340660327419 .................................................. Pipeline #26: Score on D2: 0.0018903804402523594 | D1-D2 diff: 4.969422663812882 Pipeline steps: SelectPercentile(percentile=35), HeterosisEncoder(), RecessiveEncoder(), SelectPercentile(percentile=95), SelectPercentile(percentile=45), DecisionTreeRegressor(max_depth=1, min_samples_leaf=6, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03391857272310739 Holdout data R^2 trained on entire dataset(80%): 0.04435696385660659 Dataset D1 R^2 on trained D1: 0.003530125211487656 .................................................. Pipeline #27: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), DominantEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=16, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 30 - 6 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.15938 Best score on D1-D2 diff: 5.81271 Gen 2 - Best score on D2: 0.16102 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.17046 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.17046 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.17046 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.17046 Best score on D1-D2 diff: 11.96036 Gen 7 - Best score on D2: 0.17046 Best score on D1-D2 diff: 11.96036 Gen 8 - Best score on D2: 0.17358 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.17358 Best score on D1-D2 diff: 13.48015 Gen 13 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.014297235198258429 Entire dataset(80%) R^2 trained on data (80%): 0.00608755764735891 Holdout R^2 (20%) trained on data (80%): -0.012404467813686537 Dataset D1 R^2 on trained D1: 0.013549970563374147 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0038597631789185627 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.17489283686849844 | D1-D2 diff: 1.419921605512898 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=4, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3790976486514992 Holdout data R^2 trained on entire dataset(80%): 0.22054235510434184 Dataset D1 R^2 on trained D1: 0.3796226150745243 .................................................. Pipeline #2: Score on D2: 0.1684809524019597 | D1-D2 diff: 1.4827893930200293 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=2, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3447889449965629 Holdout data R^2 trained on entire dataset(80%): 0.20736466697658917 Dataset D1 R^2 on trained D1: 0.34722872382217473 .................................................. Pipeline #3: Score on D2: 0.16712796198430568 | D1-D2 diff: 1.554705303435597 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=4, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33163271102995384 Holdout data R^2 trained on entire dataset(80%): 0.20943757376659877 Dataset D1 R^2 on trained D1: 0.3349314097702115 .................................................. Pipeline #4: Score on D2: 0.1631170435365228 | D1-D2 diff: 1.7056420254319624 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3002786763687285 Holdout data R^2 trained on entire dataset(80%): 0.2040531073225409 Dataset D1 R^2 on trained D1: 0.300314340626226 .................................................. Pipeline #5: Score on D2: 0.16045871274822243 | D1-D2 diff: 1.7173312868087474 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.2), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30283152556931325 Holdout data R^2 trained on entire dataset(80%): 0.2025259666629453 Dataset D1 R^2 on trained D1: 0.2978743474354063 .................................................. Pipeline #6: Score on D2: 0.1567023718429399 | D1-D2 diff: 1.8370626828202024 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.45, min_samples_leaf=19, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2959720695644368 Holdout data R^2 trained on entire dataset(80%): 0.20662059930987675 Dataset D1 R^2 on trained D1: 0.29223437754793535 .................................................. Pipeline #7: Score on D2: 0.12577697754328265 | D1-D2 diff: 2.496154518868005 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2870729490367444 Holdout data R^2 trained on entire dataset(80%): 0.20289843078792713 Dataset D1 R^2 on trained D1: 0.28521336444390266 .................................................. Pipeline #8: Score on D2: 0.11032356059387838 | D1-D2 diff: 2.635943056645064 Pipeline steps: SelectPercentile(percentile=80), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2767689796044962 Holdout data R^2 trained on entire dataset(80%): 0.20893917278849583 Dataset D1 R^2 on trained D1: 0.26993540683708517 .................................................. Pipeline #9: Score on D2: 0.07236622903347922 | D1-D2 diff: 2.7512071139112404 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26364754867913376 Holdout data R^2 trained on entire dataset(80%): 0.2066840994064253 Dataset D1 R^2 on trained D1: 0.25982259213765135 .................................................. Pipeline #10: Score on D2: 0.04448216570909591 | D1-D2 diff: 4.993405982569102 Pipeline steps: UnderDominanceEncoder(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), DecisionTreeRegressor(max_depth=2, min_samples_leaf=15, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2567728374422119 Holdout data R^2 trained on entire dataset(80%): 0.2040435844370483 Dataset D1 R^2 on trained D1: 0.25305229452512 .................................................. Pipeline #11: Score on D2: 0.008073274891013282 | D1-D2 diff: 6.060059571063254 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2573222725638159 Holdout data R^2 trained on entire dataset(80%): 0.20696355854746917 Dataset D1 R^2 on trained D1: 0.2512613243141134 .................................................. Pipeline #12: Score on D2: 6.368641417653365e-05 | D1-D2 diff: 11.599872543826846 Pipeline steps: SelectPercentile(percentile=5), VarianceThreshold(threshold=0.25), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25878049473603293 Holdout data R^2 trained on entire dataset(80%): 0.20767201693325577 Dataset D1 R^2 on trained D1: 0.2502170290062854 .................................................. Pipeline #13: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.8, min_samples_leaf=4, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24527834310329288 Holdout data R^2 trained on entire dataset(80%): 0.20132173473494508 Dataset D1 R^2 on trained D1: 0.23885430031740518 .................................................. ************************************************************************************** Random Seed 30 - 7 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.15938 Best score on D1-D2 diff: 5.81271 Gen 2 - Best score on D2: 0.16102 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.17046 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.17046 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.17046 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.17046 Best score on D1-D2 diff: 11.96036 Gen 7 - Best score on D2: 0.17046 Best score on D1-D2 diff: 11.96036 Gen 8 - Best score on D2: 0.17358 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.17358 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.013246410054552182 Entire dataset(80%) R^2 trained on data (80%): 0.00608755764735891 Holdout R^2 (20%) trained on data (80%): -0.012404467813686537 Dataset D1 R^2 on trained D1: 0.013549970563374147 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0038597631789185627 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.17357681583298057 | D1-D2 diff: 1.4842568393705764 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=2, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3790976486514992 Holdout data R^2 trained on entire dataset(80%): 0.22054235510434184 Dataset D1 R^2 on trained D1: 0.3796226150745243 .................................................. Pipeline #2: Score on D2: 0.16799621444152413 | D1-D2 diff: 1.536900683311958 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=10, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3447889449965629 Holdout data R^2 trained on entire dataset(80%): 0.20736466697658917 Dataset D1 R^2 on trained D1: 0.34722872382217473 .................................................. Pipeline #3: Score on D2: 0.16697846251924842 | D1-D2 diff: 1.5620793551398 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=11, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33163271102995384 Holdout data R^2 trained on entire dataset(80%): 0.20943757376659877 Dataset D1 R^2 on trained D1: 0.3349314097702115 .................................................. Pipeline #4: Score on D2: 0.16432475087592524 | D1-D2 diff: 1.6467344992048656 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=9, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3002786763687285 Holdout data R^2 trained on entire dataset(80%): 0.2040531073225409 Dataset D1 R^2 on trained D1: 0.300314340626226 .................................................. Pipeline #5: Score on D2: 0.16247099046917546 | D1-D2 diff: 1.6485140089040493 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30283152556931325 Holdout data R^2 trained on entire dataset(80%): 0.2025259666629453 Dataset D1 R^2 on trained D1: 0.2978743474354063 .................................................. Pipeline #6: Score on D2: 0.16223138727721298 | D1-D2 diff: 1.665373556751537 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=15, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2959720695644368 Holdout data R^2 trained on entire dataset(80%): 0.20662059930987675 Dataset D1 R^2 on trained D1: 0.29223437754793535 .................................................. Pipeline #7: Score on D2: 0.1610226709362852 | D1-D2 diff: 1.6845260713485077 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=16, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2870729490367444 Holdout data R^2 trained on entire dataset(80%): 0.20289843078792713 Dataset D1 R^2 on trained D1: 0.28521336444390266 .................................................. Pipeline #8: Score on D2: 0.1600959861342679 | D1-D2 diff: 1.7370424886295142 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.7000000000000001, min_samples_leaf=16, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2767689796044962 Holdout data R^2 trained on entire dataset(80%): 0.20893917278849583 Dataset D1 R^2 on trained D1: 0.26993540683708517 .................................................. Pipeline #9: Score on D2: 0.15813647780419493 | D1-D2 diff: 1.7708614625716146 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26364754867913376 Holdout data R^2 trained on entire dataset(80%): 0.2066840994064253 Dataset D1 R^2 on trained D1: 0.25982259213765135 .................................................. Pipeline #10: Score on D2: 0.15631649748116117 | D1-D2 diff: 1.7930945742503408 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2567728374422119 Holdout data R^2 trained on entire dataset(80%): 0.2040435844370483 Dataset D1 R^2 on trained D1: 0.25305229452512 .................................................. Pipeline #11: Score on D2: 0.15625524949868763 | D1-D2 diff: 1.8012009034008067 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.05), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2573222725638159 Holdout data R^2 trained on entire dataset(80%): 0.20696355854746917 Dataset D1 R^2 on trained D1: 0.2512613243141134 .................................................. Pipeline #12: Score on D2: 0.1558624295059733 | D1-D2 diff: 1.8043020024522443 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25878049473603293 Holdout data R^2 trained on entire dataset(80%): 0.20767201693325577 Dataset D1 R^2 on trained D1: 0.2502170290062854 .................................................. Pipeline #13: Score on D2: 0.15431676514091297 | D1-D2 diff: 1.8545459774201436 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24527834310329288 Holdout data R^2 trained on entire dataset(80%): 0.20132173473494508 Dataset D1 R^2 on trained D1: 0.23885430031740518 .................................................. Pipeline #14: Score on D2: 0.15244336736273545 | D1-D2 diff: 1.8727628932586893 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24197116490302606 Holdout data R^2 trained on entire dataset(80%): 0.19676158214741035 Dataset D1 R^2 on trained D1: 0.23373929990148923 .................................................. Pipeline #15: Score on D2: 0.12265516180716984 | D1-D2 diff: 2.393808890756757 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=6, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14017853754749143 Holdout data R^2 trained on entire dataset(80%): 0.1780490026783187 Dataset D1 R^2 on trained D1: 0.15310900431337404 .................................................. Pipeline #16: Score on D2: 0.07937216544435222 | D1-D2 diff: 2.446325952241003 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), DecisionTreeRegressor(max_depth=3, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0943408645030317 Holdout data R^2 trained on entire dataset(80%): 0.1264888918941921 Dataset D1 R^2 on trained D1: 0.10729392028132789 .................................................. Pipeline #17: Score on D2: 0.07278109940046396 | D1-D2 diff: 2.781043837146251 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), DecisionTreeRegressor(max_depth=2, min_samples_leaf=6, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08159293235943832 Holdout data R^2 trained on entire dataset(80%): 0.12318927968790461 Dataset D1 R^2 on trained D1: 0.08949849660781306 .................................................. Pipeline #18: Score on D2: 0.038973936919545316 | D1-D2 diff: 2.789150419079063 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04778629510930643 Holdout data R^2 trained on entire dataset(80%): 0.04743307387056406 Dataset D1 R^2 on trained D1: 0.055497825362541664 .................................................. Pipeline #19: Score on D2: 0.03731354559314304 | D1-D2 diff: 2.943871899462998 Pipeline steps: VarianceThreshold(threshold=0.15), DecisionTreeRegressor(max_depth=2, min_samples_leaf=6, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.044364755061259586 Holdout data R^2 trained on entire dataset(80%): 0.04914508831215614 Dataset D1 R^2 on trained D1: 0.05062803044148678 .................................................. Pipeline #20: Score on D2: 0.036705996583596856 | D1-D2 diff: 2.973223444694779 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04345804458561919 Holdout data R^2 trained on entire dataset(80%): 0.0364356432513675 Dataset D1 R^2 on trained D1: 0.04950245545869636 .................................................. Pipeline #21: Score on D2: 0.027552520777431422 | D1-D2 diff: 2.985636788420336 Pipeline steps: SelectPercentile(percentile=25), DecisionTreeRegressor(max_depth=2, min_samples_leaf=6, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03571048735858029 Holdout data R^2 trained on entire dataset(80%): 0.024375239840601526 Dataset D1 R^2 on trained D1: 0.04013748851212995 .................................................. Pipeline #22: Score on D2: 0.02604117358576785 | D1-D2 diff: 3.2093089368239474 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.5, min_samples_leaf=19, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.031034173869150194 Holdout data R^2 trained on entire dataset(80%): 0.031813688121002714 Dataset D1 R^2 on trained D1: 0.0354677479533424 .................................................. Pipeline #23: Score on D2: 0.022319431676171897 | D1-D2 diff: 4.30547282696919 Pipeline steps: OverDominanceEncoder(), HeterosisEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.023862587037685512 Holdout data R^2 trained on entire dataset(80%): 0.009062281979811626 Dataset D1 R^2 on trained D1: 0.025229589816299747 .................................................. Pipeline #24: Score on D2: 0.02230167103805103 | D1-D2 diff: 5.163380743525418 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.023190468245248286 Holdout data R^2 trained on entire dataset(80%): 0.016784186318788752 Dataset D1 R^2 on trained D1: 0.023708571580195126 .................................................. Pipeline #25: Score on D2: 0.012238487659723463 | D1-D2 diff: 5.812710923972204 Pipeline steps: DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.012713437088444413 Holdout data R^2 trained on entire dataset(80%): 0.0071001017078963224 Dataset D1 R^2 on trained D1: 0.013114449321569799 .................................................. Pipeline #26: Score on D2: 0.012237095314328261 | D1-D2 diff: 5.822752943573516 Pipeline steps: RecessiveEncoder(), SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.05, min_samples_leaf=11, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01271227176183376 Holdout data R^2 trained on entire dataset(80%): 0.007072759749748858 Dataset D1 R^2 on trained D1: 0.013107029796440739 .................................................. Pipeline #27: Score on D2: 0.003089017253358861 | D1-D2 diff: 6.270261020617466 Pipeline steps: SelectPercentile(percentile=5), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0028048970030285503 Holdout data R^2 trained on entire dataset(80%): -0.005703644074040559 Dataset D1 R^2 on trained D1: 0.0024420869133796597 .................................................. Pipeline #28: Score on D2: 0.0030470601975841616 | D1-D2 diff: 7.872976585258862 Pipeline steps: SelectPercentile(percentile=10), RecessiveEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002947519786774211 Holdout data R^2 trained on entire dataset(80%): -0.004461250201079459 Dataset D1 R^2 on trained D1: 0.002786778182288452 .................................................. Pipeline #29: Score on D2: 0.0030470601975839395 | D1-D2 diff: 7.872976585260542 Pipeline steps: SelectPercentile(percentile=20), RecessiveEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002947519786774211 Holdout data R^2 trained on entire dataset(80%): -0.004461250201079459 Dataset D1 R^2 on trained D1: 0.002786778182288452 .................................................. Pipeline #30: Score on D2: 0.0029852546004652813 | D1-D2 diff: 8.535029125782614 Pipeline steps: SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.05, min_samples_leaf=11, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0029566645365888267 Holdout data R^2 trained on entire dataset(80%): -0.00483460429371263 Dataset D1 R^2 on trained D1: 0.002796811617601258 .................................................. Pipeline #31: Score on D2: 0.002959776265629044 | D1-D2 diff: 8.893587579509557 Pipeline steps: SelectPercentile(percentile=5), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.05, min_samples_leaf=11, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002958849626700566 Holdout data R^2 trained on entire dataset(80%): -0.004782723630507135 Dataset D1 R^2 on trained D1: 0.0027999338421839104 .................................................. Pipeline #32: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.8, min_samples_leaf=9, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 30 - 8 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.17670 Best score on D1-D2 diff: 3.57196 Gen 2 - Best score on D2: 0.17670 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.17670 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.18554 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.18554 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.18554 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.18554 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.19221 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.19221 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.19221 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.19221 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.19221 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.19221 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.19379 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0068766693969233295 Entire dataset(80%) R^2 trained on data (80%): 0.007359247303647343 Holdout R^2 (20%) trained on data (80%): -0.005580483630530209 Dataset D1 R^2 on trained D1: 0.014812199294860329 Combined Dataset (100%) R^2 trained on combined data (100%): 0.005857247210122152 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.19379366690394462 | D1-D2 diff: 1.4720026659522885 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4072416039189588 Holdout data R^2 trained on entire dataset(80%): 0.22455681807973427 Dataset D1 R^2 on trained D1: 0.406786805776866 .................................................. Pipeline #2: Score on D2: 0.19192436298888504 | D1-D2 diff: 1.5727507435256676 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), OverDominanceEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3600673829955957 Holdout data R^2 trained on entire dataset(80%): 0.22714033918017107 Dataset D1 R^2 on trained D1: 0.3553651343589579 .................................................. Pipeline #3: Score on D2: 0.18913822488139742 | D1-D2 diff: 1.5803189660134744 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3539220888201484 Holdout data R^2 trained on entire dataset(80%): 0.22563636394535114 Dataset D1 R^2 on trained D1: 0.34947051316418254 .................................................. Pipeline #4: Score on D2: 0.1885346144306992 | D1-D2 diff: 1.5858069398175079 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), OverDominanceEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.35379491192941703 Holdout data R^2 trained on entire dataset(80%): 0.2274267051767931 Dataset D1 R^2 on trained D1: 0.34665896093530346 .................................................. Pipeline #5: Score on D2: 0.1827705892457332 | D1-D2 diff: 1.6136254750274794 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), VarianceThreshold(threshold=0.15), UnderDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3386977878094749 Holdout data R^2 trained on entire dataset(80%): 0.22891888485819556 Dataset D1 R^2 on trained D1: 0.3302695747858597 .................................................. Pipeline #6: Score on D2: 0.18120555168504193 | D1-D2 diff: 1.6849728871709984 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=15, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30911347879752316 Holdout data R^2 trained on entire dataset(80%): 0.2157328771974487 Dataset D1 R^2 on trained D1: 0.3052645675875244 .................................................. Pipeline #7: Score on D2: 0.17938328389502722 | D1-D2 diff: 1.7150066899067615 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=15, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29858433746982105 Holdout data R^2 trained on entire dataset(80%): 0.21225658744414233 Dataset D1 R^2 on trained D1: 0.29497766552726035 .................................................. Pipeline #8: Score on D2: 0.17893015800401424 | D1-D2 diff: 1.7312477254597183 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29624657658192155 Holdout data R^2 trained on entire dataset(80%): 0.2130069604775 Dataset D1 R^2 on trained D1: 0.2902475791690924 .................................................. Pipeline #9: Score on D2: 0.1775006613055845 | D1-D2 diff: 1.737835048700148 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2923729446807368 Holdout data R^2 trained on entire dataset(80%): 0.21282559755064367 Dataset D1 R^2 on trained D1: 0.2871398448133454 .................................................. Pipeline #10: Score on D2: 0.1767687122611583 | D1-D2 diff: 1.7479967512750452 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2857719122488033 Holdout data R^2 trained on entire dataset(80%): 0.20986784825640925 Dataset D1 R^2 on trained D1: 0.2838805610420223 .................................................. Pipeline #11: Score on D2: 0.17669598819881094 | D1-D2 diff: 1.749023534310878 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28642758002594915 Holdout data R^2 trained on entire dataset(80%): 0.21266964832452206 Dataset D1 R^2 on trained D1: 0.2835565337420811 .................................................. Pipeline #12: Score on D2: 0.17633110276760344 | D1-D2 diff: 1.7900314386983511 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27686758080340246 Holdout data R^2 trained on entire dataset(80%): 0.21024626376957956 Dataset D1 R^2 on trained D1: 0.2737307458425886 .................................................. Pipeline #13: Score on D2: 0.17480236467726196 | D1-D2 diff: 1.8224044902638612 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=20, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27105783239703385 Holdout data R^2 trained on entire dataset(80%): 0.20853750074541244 Dataset D1 R^2 on trained D1: 0.26546344561592483 .................................................. Pipeline #14: Score on D2: 0.17043799054181663 | D1-D2 diff: 1.8350076272584863 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=15, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2629645459220773 Holdout data R^2 trained on entire dataset(80%): 0.21007069624555963 Dataset D1 R^2 on trained D1: 0.2586339126033639 .................................................. Pipeline #15: Score on D2: 0.16891796500842482 | D1-D2 diff: 1.868833079598114 Pipeline steps: UnderDominanceEncoder(), UnderDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2594930199683564 Holdout data R^2 trained on entire dataset(80%): 0.20980627071542413 Dataset D1 R^2 on trained D1: 0.25089985925946634 .................................................. Pipeline #16: Score on D2: 0.1679863052276983 | D1-D2 diff: 1.8858629204985446 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25458170927467594 Holdout data R^2 trained on entire dataset(80%): 0.2083065788648143 Dataset D1 R^2 on trained D1: 0.2470467974160384 .................................................. Pipeline #17: Score on D2: 0.16767767894793983 | D1-D2 diff: 1.8996422647834097 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=14, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25175285659524904 Holdout data R^2 trained on entire dataset(80%): 0.2043527127216681 Dataset D1 R^2 on trained D1: 0.2444691002388495 .................................................. Pipeline #18: Score on D2: 0.16613049940786984 | D1-D2 diff: 1.900980151734596 Pipeline steps: UnderDominanceEncoder(), UnderDominanceEncoder(), DominantEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2502877610726727 Holdout data R^2 trained on entire dataset(80%): 0.20740369675787562 Dataset D1 R^2 on trained D1: 0.24270596929626143 .................................................. Pipeline #19: Score on D2: 0.16589144577633608 | D1-D2 diff: 1.9042612216804775 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24695548495108977 Holdout data R^2 trained on entire dataset(80%): 0.20555224883371404 Dataset D1 R^2 on trained D1: 0.2419405154927432 .................................................. Pipeline #20: Score on D2: 0.16588741622343295 | D1-D2 diff: 1.9280780150902441 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2466299341727416 Holdout data R^2 trained on entire dataset(80%): 0.20312283631341255 Dataset D1 R^2 on trained D1: 0.23824792129383854 .................................................. Pipeline #21: Score on D2: 0.16497150063892663 | D1-D2 diff: 1.9785755158185399 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23587142261667005 Holdout data R^2 trained on entire dataset(80%): 0.20131575032921512 Dataset D1 R^2 on trained D1: 0.2302228471158717 .................................................. Pipeline #22: Score on D2: 0.16254642116928697 | D1-D2 diff: 2.0127536729267472 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=18, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2346384459862163 Holdout data R^2 trained on entire dataset(80%): 0.19987759517290504 Dataset D1 R^2 on trained D1: 0.2234773065261677 .................................................. Pipeline #23: Score on D2: 0.15538637895970808 | D1-D2 diff: 2.0257631081342287 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=17, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22285417201285507 Holdout data R^2 trained on entire dataset(80%): 0.19176460514197413 Dataset D1 R^2 on trained D1: 0.21476708661147415 .................................................. Pipeline #24: Score on D2: 0.15248591609501572 | D1-D2 diff: 2.0284379809088247 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=17, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2236006378237131 Holdout data R^2 trained on entire dataset(80%): 0.18897744766622626 Dataset D1 R^2 on trained D1: 0.211554024724392 .................................................. Pipeline #25: Score on D2: 0.1465158036961718 | D1-D2 diff: 2.0363480548527186 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=15, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2123010580876431 Holdout data R^2 trained on entire dataset(80%): 0.18372420782116639 Dataset D1 R^2 on trained D1: 0.20467145978114842 .................................................. Pipeline #26: Score on D2: 0.14471423374853298 | D1-D2 diff: 2.0976339610998 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=16, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2081585397205452 Holdout data R^2 trained on entire dataset(80%): 0.1825094360697782 Dataset D1 R^2 on trained D1: 0.19636552429201526 .................................................. Pipeline #27: Score on D2: 0.12685991629363946 | D1-D2 diff: 2.1425260519579425 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), VarianceThreshold(), RandomForestRegressor(bootstrap=False, max_features=0.1, min_samples_leaf=20, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19208819711030745 Holdout data R^2 trained on entire dataset(80%): 0.15995304009921207 Dataset D1 R^2 on trained D1: 0.17431639982221336 .................................................. Pipeline #28: Score on D2: 0.11944814008433302 | D1-D2 diff: 2.4571454171229665 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(threshold=0.15), FeatureEncodingFrequencySelector(threshold=0.25), DecisionTreeRegressor(max_depth=4, min_samples_leaf=18, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14754744499936068 Holdout data R^2 trained on entire dataset(80%): 0.15697559165283081 Dataset D1 R^2 on trained D1: 0.1468813459490762 .................................................. Pipeline #29: Score on D2: 0.11894247161781657 | D1-D2 diff: 2.475745030267012 Pipeline steps: VarianceThreshold(threshold=0.1), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14728718842108834 Holdout data R^2 trained on entire dataset(80%): 0.16976416023216268 Dataset D1 R^2 on trained D1: 0.14556052773147476 .................................................. Pipeline #30: Score on D2: 0.11268831106282029 | D1-D2 diff: 2.8709642840993688 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), DecisionTreeRegressor(max_depth=3, min_samples_leaf=8, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12055391119417447 Holdout data R^2 trained on entire dataset(80%): 0.13924168634835332 Dataset D1 R^2 on trained D1: 0.12740766814193605 .................................................. Pipeline #31: Score on D2: 0.046566687375021565 | D1-D2 diff: 3.644555736383321 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04951729070213762 Holdout data R^2 trained on entire dataset(80%): 0.0621036591196078 Dataset D1 R^2 on trained D1: 0.05223458000776182 .................................................. Pipeline #32: Score on D2: 0.04656651537917278 | D1-D2 diff: 3.6487534870472804 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0495015720963311 Holdout data R^2 trained on entire dataset(80%): 0.062028481359970744 Dataset D1 R^2 on trained D1: 0.05220837021761926 .................................................. Pipeline #33: Score on D2: 0.04494491733397943 | D1-D2 diff: 4.220573623171039 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04704944966854896 Holdout data R^2 trained on entire dataset(80%): 0.05476305986795671 Dataset D1 R^2 on trained D1: 0.048096393871540966 .................................................. Pipeline #34: Score on D2: 0.04004821777398648 | D1-D2 diff: 4.9338127407385866 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.039223366074151333 Holdout data R^2 trained on entire dataset(80%): 0.055327711125045975 Dataset D1 R^2 on trained D1: 0.03836061841614369 .................................................. Pipeline #35: Score on D2: 0.00592085568156242 | D1-D2 diff: 6.471862213552641 Pipeline steps: SelectPercentile(percentile=10), UnderDominanceEncoder(), UnderDominanceEncoder(), DominantEncoder(), SelectPercentile(percentile=25), OverDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.005645418131205671 Holdout data R^2 trained on entire dataset(80%): 0.015639494884561955 Dataset D1 R^2 on trained D1: 0.005350845058080478 .................................................. Pipeline #36: Score on D2: 0.005899987899755876 | D1-D2 diff: 6.5411635044437055 Pipeline steps: SelectPercentile(percentile=5), VarianceThreshold(threshold=0.05), VarianceThreshold(threshold=0.1), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.005647669737357974 Holdout data R^2 trained on entire dataset(80%): 0.01588059116454299 Dataset D1 R^2 on trained D1: 0.005353752326323624 .................................................. Pipeline #37: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), SelectPercentile(percentile=85), RandomForestRegressor(max_features=1.0, min_samples_leaf=17, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 30 - 9 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.18172 Best score on D1-D2 diff: 3.64553 Gen 2 - Best score on D2: 0.18172 Best score on D1-D2 diff: 4.34662 Gen 3 - Best score on D2: 0.18199 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.18199 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.18540 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.18546 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.18546 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.18546 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.18546 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.18546 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.18546 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.18546 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.18546 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.18831 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.18831 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.18831 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.18831 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.19168 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.19168 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.19168 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.19391 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.19391 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.19391 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.19391 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.19391 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0015372477882800162 Entire dataset(80%) R^2 trained on data (80%): 0.007521703151726178 Holdout R^2 (20%) trained on data (80%): 0.0010819069258051206 Dataset D1 R^2 on trained D1: 0.011511245861252517 Combined Dataset (100%) R^2 trained on combined data (100%): 0.006896448179604553 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.19390959307004585 | D1-D2 diff: 1.4201933465665202 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=4, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4243674743520144 Holdout data R^2 trained on entire dataset(80%): 0.23626870097830766 Dataset D1 R^2 on trained D1: 0.4397255691777011 .................................................. Pipeline #2: Score on D2: 0.19169811531999437 | D1-D2 diff: 1.4807500624276173 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), VarianceThreshold(threshold=0.1), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=3, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3902487606220061 Holdout data R^2 trained on entire dataset(80%): 0.23860206540992002 Dataset D1 R^2 on trained D1: 0.39970272588728706 .................................................. Pipeline #3: Score on D2: 0.1890298740926497 | D1-D2 diff: 1.4996873287176071 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=4, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3748148445727728 Holdout data R^2 trained on entire dataset(80%): 0.23581049449952518 Dataset D1 R^2 on trained D1: 0.38672552343018696 .................................................. Pipeline #4: Score on D2: 0.188305236929741 | D1-D2 diff: 1.523479592408117 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=4, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3670581348902573 Holdout data R^2 trained on entire dataset(80%): 0.23437605500551795 Dataset D1 R^2 on trained D1: 0.3739374899639901 .................................................. Pipeline #5: Score on D2: 0.1876664428546102 | D1-D2 diff: 1.5357117689831439 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=3, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3630927884912202 Holdout data R^2 trained on entire dataset(80%): 0.23454791810966602 Dataset D1 R^2 on trained D1: 0.36745462859973954 .................................................. Pipeline #6: Score on D2: 0.1868370112133909 | D1-D2 diff: 1.563374012887144 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=2, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.35304806776783537 Holdout data R^2 trained on entire dataset(80%): 0.23364705858663892 Dataset D1 R^2 on trained D1: 0.3542343098971017 .................................................. Pipeline #7: Score on D2: 0.18575785449393178 | D1-D2 diff: 1.599788282918884 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=4, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3347459859288434 Holdout data R^2 trained on entire dataset(80%): 0.23343921156033554 Dataset D1 R^2 on trained D1: 0.3384265355003049 .................................................. Pipeline #8: Score on D2: 0.1850233478897224 | D1-D2 diff: 1.6921045434259825 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31272525898477843 Holdout data R^2 trained on entire dataset(80%): 0.2283827832094344 Dataset D1 R^2 on trained D1: 0.3070040794588511 .................................................. Pipeline #9: Score on D2: 0.18171885346366157 | D1-D2 diff: 1.74961551333072 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2887322059952472 Holdout data R^2 trained on entire dataset(80%): 0.23169286182288273 Dataset D1 R^2 on trained D1: 0.2884348481560097 .................................................. Pipeline #10: Score on D2: 0.1788377627436899 | D1-D2 diff: 1.7638450331858895 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=10, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2806159109155366 Holdout data R^2 trained on entire dataset(80%): 0.22770406280423172 Dataset D1 R^2 on trained D1: 0.28215155283249327 .................................................. Pipeline #11: Score on D2: 0.1760608515854838 | D1-D2 diff: 1.7653996124717142 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), VarianceThreshold(threshold=0.1), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2806515250433863 Holdout data R^2 trained on entire dataset(80%): 0.22512293670347228 Dataset D1 R^2 on trained D1: 0.27901121697056197 .................................................. Pipeline #12: Score on D2: 0.17574033361211983 | D1-D2 diff: 1.7904722192387181 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27670980564348935 Holdout data R^2 trained on entire dataset(80%): 0.22229109909672518 Dataset D1 R^2 on trained D1: 0.27304410026764214 .................................................. Pipeline #13: Score on D2: 0.17389735483907343 | D1-D2 diff: 1.8000273478842517 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2713049475395165 Holdout data R^2 trained on entire dataset(80%): 0.23449319536280278 Dataset D1 R^2 on trained D1: 0.2691514347461159 .................................................. Pipeline #14: Score on D2: 0.17371765144321305 | D1-D2 diff: 1.8689966519838588 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=13, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2607218381955102 Holdout data R^2 trained on entire dataset(80%): 0.22155885395372554 Dataset D1 R^2 on trained D1: 0.2556708496263803 .................................................. Pipeline #15: Score on D2: 0.1729973868775292 | D1-D2 diff: 1.8748577688234145 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26237600524971605 Holdout data R^2 trained on entire dataset(80%): 0.22947359381950228 Dataset D1 R^2 on trained D1: 0.2539305833360793 .................................................. Pipeline #16: Score on D2: 0.17236890010403694 | D1-D2 diff: 1.8969981884725964 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=16, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2552267482431363 Holdout data R^2 trained on entire dataset(80%): 0.22370292623451637 Dataset D1 R^2 on trained D1: 0.24958935138749005 .................................................. Pipeline #17: Score on D2: 0.1717226298087814 | D1-D2 diff: 1.998491718232284 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24438378741728428 Holdout data R^2 trained on entire dataset(80%): 0.21425436914157625 Dataset D1 R^2 on trained D1: 0.23441152102187646 .................................................. Pipeline #18: Score on D2: 0.16627011152451665 | D1-D2 diff: 2.0310468083987723 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23526652581983776 Holdout data R^2 trained on entire dataset(80%): 0.2146892799862522 Dataset D1 R^2 on trained D1: 0.22503531850586622 .................................................. Pipeline #19: Score on D2: 0.16393724084903394 | D1-D2 diff: 2.0664147044494374 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2292459020091253 Holdout data R^2 trained on entire dataset(80%): 0.21035327293082284 Dataset D1 R^2 on trained D1: 0.21878135999739956 .................................................. Pipeline #20: Score on D2: 0.1580056590742127 | D1-D2 diff: 2.087662680591498 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21648977776524825 Holdout data R^2 trained on entire dataset(80%): 0.20005260644831402 Dataset D1 R^2 on trained D1: 0.2106508480468653 .................................................. Pipeline #21: Score on D2: 0.14602154304101045 | D1-D2 diff: 2.1433103619227123 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.1, min_samples_leaf=13, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20892524733963058 Holdout data R^2 trained on entire dataset(80%): 0.1816081581324993 Dataset D1 R^2 on trained D1: 0.1934086009406084 .................................................. Pipeline #22: Score on D2: 0.14374373072097235 | D1-D2 diff: 2.1783868405588946 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20493540502593066 Holdout data R^2 trained on entire dataset(80%): 0.18974515203927533 Dataset D1 R^2 on trained D1: 0.18815160458714553 .................................................. Pipeline #23: Score on D2: 0.13834880090595414 | D1-D2 diff: 2.1961226314572957 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19515203009670667 Holdout data R^2 trained on entire dataset(80%): 0.1733716981811665 Dataset D1 R^2 on trained D1: 0.18133941513872942 .................................................. Pipeline #24: Score on D2: 0.13608419835922492 | D1-D2 diff: 2.2543248005765335 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18727723898390525 Holdout data R^2 trained on entire dataset(80%): 0.17510853192668907 Dataset D1 R^2 on trained D1: 0.17480408208673148 .................................................. Pipeline #25: Score on D2: 0.1343269604968338 | D1-D2 diff: 2.254672659271653 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19158069514166276 Holdout data R^2 trained on entire dataset(80%): 0.1757251310256055 Dataset D1 R^2 on trained D1: 0.1730229544034977 .................................................. Pipeline #26: Score on D2: 0.1328274179662009 | D1-D2 diff: 2.293017296629322 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.3), HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.183793406606385 Holdout data R^2 trained on entire dataset(80%): 0.16985985805377757 Dataset D1 R^2 on trained D1: 0.1689992643002578 .................................................. Pipeline #27: Score on D2: 0.13161460482437803 | D1-D2 diff: 2.3107392136983256 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.3), HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1800882630791356 Holdout data R^2 trained on entire dataset(80%): 0.16689279616915953 Dataset D1 R^2 on trained D1: 0.16668949033635094 .................................................. Pipeline #28: Score on D2: 0.13034014600220178 | D1-D2 diff: 2.3152185671582446 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.3), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.181839444839345 Holdout data R^2 trained on entire dataset(80%): 0.16634381830710543 Dataset D1 R^2 on trained D1: 0.16514437466923926 .................................................. Pipeline #29: Score on D2: 0.1275627583838398 | D1-D2 diff: 2.6457786303858497 Pipeline steps: OverDominanceEncoder(), OverDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=9, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1294654275789472 Holdout data R^2 trained on entire dataset(80%): 0.17195127241114283 Dataset D1 R^2 on trained D1: 0.14797007875393642 .................................................. Pipeline #30: Score on D2: 0.09922104552343392 | D1-D2 diff: 2.7165611710504836 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0906447372355138 Holdout data R^2 trained on entire dataset(80%): 0.1307904755398559 Dataset D1 R^2 on trained D1: 0.08085895836052737 .................................................. Pipeline #31: Score on D2: 0.09742435208914535 | D1-D2 diff: 3.4891894187289614 Pipeline steps: SelectPercentile(percentile=75), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=17, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10494362494135312 Holdout data R^2 trained on entire dataset(80%): 0.14577166758629878 Dataset D1 R^2 on trained D1: 0.10417121386719708 .................................................. Pipeline #32: Score on D2: 0.07619840742741102 | D1-D2 diff: 3.975148755136361 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0790544427579839 Holdout data R^2 trained on entire dataset(80%): 0.10774340820108008 Dataset D1 R^2 on trained D1: 0.08020325931777728 .................................................. Pipeline #33: Score on D2: 0.06829091261774856 | D1-D2 diff: 4.211960667432865 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), VarianceThreshold(threshold=0.25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06695796461054226 Holdout data R^2 trained on entire dataset(80%): 0.0967381119402212 Dataset D1 R^2 on trained D1: 0.06511357933408801 .................................................. Pipeline #34: Score on D2: 0.044944917392994777 | D1-D2 diff: 4.220573667196048 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=20, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05975088737122547 Holdout data R^2 trained on entire dataset(80%): 0.08335524848710474 Dataset D1 R^2 on trained D1: 0.0480963937990635 .................................................. Pipeline #35: Score on D2: 0.043118227133936426 | D1-D2 diff: 12.416520332080998 Pipeline steps: DominantEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04350600027698248 Holdout data R^2 trained on entire dataset(80%): 0.04048778925206198 Dataset D1 R^2 on trained D1: 0.04307615443390855 .................................................. Pipeline #36: Score on D2: 0.043118227133936204 | D1-D2 diff: 12.416520332097381 Pipeline steps: DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04350600027698226 Holdout data R^2 trained on entire dataset(80%): 0.04048778925206198 Dataset D1 R^2 on trained D1: 0.04307615443390855 .................................................. Pipeline #37: Score on D2: -3.2015594250811574e-05 | D1-D2 diff: 13.480151627526027 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021467281316e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516763899763 Dataset D1 R^2 on trained D1: -1.73106056000627e-06 .................................................. ************************************************************************************** Random Seed 31 - 0 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00181 Best score on D1-D2 diff: 3.79472 Gen 2 - Best score on D2: -0.00170 Best score on D1-D2 diff: 4.25082 Gen 3 - Best score on D2: 0.00018 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.00089 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.00089 Best score on D1-D2 diff: 11.96036 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0097466550613734 Entire dataset(80%) R^2 trained on data (80%): 0.0024885120238450353 Holdout R^2 (20%) trained on data (80%): -0.007051783125658018 Dataset D1 R^2 on trained D1: 0.006635562118047389 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0014844060778657076 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.0008896100496513792 | D1-D2 diff: 3.768990715050282 Pipeline steps: VarianceThreshold(threshold=0.05), RecessiveEncoder(), SelectPercentile(percentile=20), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004044653158699951 Holdout data R^2 trained on entire dataset(80%): -0.002966515325050212 Dataset D1 R^2 on trained D1: 0.005845249821544551 .................................................. Pipeline #2: Score on D2: 0.0003442878604913435 | D1-D2 diff: 3.800852969502792 Pipeline steps: RecessiveEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0033975880942680448 Holdout data R^2 trained on entire dataset(80%): -0.0030587029054938153 Dataset D1 R^2 on trained D1: 0.005135834507749348 .................................................. Pipeline #3: Score on D2: 0.00019887049380440391 | D1-D2 diff: 5.731567223265859 Pipeline steps: RecessiveEncoder(), VarianceThreshold(threshold=0.15), SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.8, min_samples_leaf=17, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0007685079279374785 Holdout data R^2 trained on entire dataset(80%): -0.0016070028991348462 Dataset D1 R^2 on trained D1: 0.0011255006681076818 .................................................. Pipeline #4: Score on D2: -4.886781085700065e-05 | D1-D2 diff: 11.960360745435652 Pipeline steps: RecessiveEncoder(), DominantEncoder(), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0 Holdout data R^2 trained on entire dataset(80%): -0.0012028184018280097 Dataset D1 R^2 on trained D1: 0.0 .................................................. ************************************************************************************** Random Seed 31 - 1 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.05294 Best score on D1-D2 diff: 3.79472 Gen 2 - Best score on D2: 0.05294 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.05294 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.06314 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.07238 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.07238 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.07238 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.07238 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.07238 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.07238 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.07852 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.010626275504642901 Entire dataset(80%) R^2 trained on data (80%): 0.002196933490644204 Holdout R^2 (20%) trained on data (80%): -0.006030951706914811 Dataset D1 R^2 on trained D1: 0.006562934865342651 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0013773220823013466 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.07851861790147074 | D1-D2 diff: 2.025020365189038 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), SelectPercentile(percentile=35), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=17, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.013067707494140546 Holdout data R^2 trained on entire dataset(80%): 0.00017936913052163383 Dataset D1 R^2 on trained D1: 0.13798649282050035 .................................................. Pipeline #2: Score on D2: 0.06635823394314777 | D1-D2 diff: 2.0608660443900013 Pipeline steps: VarianceThreshold(threshold=0.25), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), UnderDominanceEncoder(), SelectPercentile(percentile=55), RandomForestRegressor(max_features=0.05, min_samples_leaf=19, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12387518648706963 Holdout data R^2 trained on entire dataset(80%): 0.08509008726809797 Dataset D1 R^2 on trained D1: 0.12179539031509445 .................................................. Pipeline #3: Score on D2: 0.05867645952472689 | D1-D2 diff: 2.0835672159986136 Pipeline steps: VarianceThreshold(threshold=0.25), FeatureEncodingFrequencySelector(threshold=0.3), UnderDominanceEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=55), RandomForestRegressor(max_features=0.05, min_samples_leaf=19, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1161092693129745 Holdout data R^2 trained on entire dataset(80%): 0.0827074873141419 Dataset D1 R^2 on trained D1: 0.11173678852102875 .................................................. Pipeline #4: Score on D2: 0.05755148078503225 | D1-D2 diff: 2.0896144329469717 Pipeline steps: VarianceThreshold(threshold=0.25), HeterosisEncoder(), UnderDominanceEncoder(), SelectPercentile(percentile=55), RandomForestRegressor(max_features=0.05, min_samples_leaf=19, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11447751599852218 Holdout data R^2 trained on entire dataset(80%): 0.0789374549484031 Dataset D1 R^2 on trained D1: 0.11000025743885378 .................................................. Pipeline #5: Score on D2: 0.056409822895006845 | D1-D2 diff: 2.377921923480965 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), SelectPercentile(percentile=35), DecisionTreeRegressor(max_depth=4, min_samples_leaf=9, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00671603246974517 Holdout data R^2 trained on entire dataset(80%): -0.005280888398496053 Dataset D1 R^2 on trained D1: 0.08768570996396963 .................................................. Pipeline #6: Score on D2: 0.0031858491150413837 | D1-D2 diff: 2.600264211860734 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=55), DominantEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=10, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.020961079040574537 Holdout data R^2 trained on entire dataset(80%): -0.003698729205269924 Dataset D1 R^2 on trained D1: 0.025059943672729457 .................................................. Pipeline #7: Score on D2: 0.0020841147091051893 | D1-D2 diff: 2.8760446349284083 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=55), DominantEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=19, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.012923212446441235 Holdout data R^2 trained on entire dataset(80%): -0.0010790943650957896 Dataset D1 R^2 on trained D1: 0.016699743785686283 .................................................. Pipeline #8: Score on D2: 0.0011887823069547387 | D1-D2 diff: 4.184571348927761 Pipeline steps: RecessiveEncoder(), SelectPercentile(percentile=30), SelectPercentile(percentile=80), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=8, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.005244617023342446 Holdout data R^2 trained on entire dataset(80%): -0.00398758403860211 Dataset D1 R^2 on trained D1: 0.004450122413504287 .................................................. Pipeline #9: Score on D2: 0.0011014169938132756 | D1-D2 diff: 4.223039384311976 Pipeline steps: RecessiveEncoder(), SelectPercentile(percentile=30), SelectPercentile(percentile=80), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=8, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.005137319339088786 Holdout data R^2 trained on entire dataset(80%): -0.004078905849411063 Dataset D1 R^2 on trained D1: 0.004245539600257753 .................................................. Pipeline #10: Score on D2: 0.0004992261421856714 | D1-D2 diff: 7.129051713575257 Pipeline steps: SelectPercentile(percentile=10), FeatureEncodingFrequencySelector(threshold=0.0), RecessiveEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=16, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0009629028116384664 Holdout data R^2 trained on entire dataset(80%): -0.0028104079965849404 Dataset D1 R^2 on trained D1: 0.0008863705048588466 .................................................. Pipeline #11: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=17, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 31 - 2 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.07493 Best score on D1-D2 diff: 3.79472 Gen 2 - Best score on D2: 0.08509 Best score on D1-D2 diff: 13.48015 Gen 3 - Best score on D2: 0.08509 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.08509 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.08509 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.08509 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.08509 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.08509 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.08943 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.08943 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.08943 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.08943 Best score on D1-D2 diff: 14.08552 Gen 13 - Best score on D2: 0.08943 Best score on D1-D2 diff: 14.08552 Gen 14 - Best score on D2: 0.09014 Best score on D1-D2 diff: 14.08552 Gen 15 - Best score on D2: 0.09014 Best score on D1-D2 diff: 14.08552 Gen 16 - Best score on D2: 0.09014 Best score on D1-D2 diff: 14.08552 Gen 17 - Best score on D2: 0.09014 Best score on D1-D2 diff: 14.08552 Gen 18 - Best score on D2: 0.09014 Best score on D1-D2 diff: 14.08552 Gen 19 - Best score on D2: 0.09014 Best score on D1-D2 diff: 14.08552 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.013310797789139572 Entire dataset(80%) R^2 trained on data (80%): 0.0023090899418051203 Holdout R^2 (20%) trained on data (80%): -0.007362308639593973 Dataset D1 R^2 on trained D1: 0.006307383295157676 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0012531315670066823 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.09013722388866052 | D1-D2 diff: 1.7117914947072315 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), RecessiveEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.201211883954988 Holdout data R^2 trained on entire dataset(80%): 0.11022277136635805 Dataset D1 R^2 on trained D1: 0.20660252215449337 .................................................. Pipeline #2: Score on D2: 0.08422912118702841 | D1-D2 diff: 1.811162752199473 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.4, min_samples_leaf=20, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17621613626115507 Holdout data R^2 trained on entire dataset(80%): 0.10864722574927066 Dataset D1 R^2 on trained D1: 0.17716214919270956 .................................................. Pipeline #3: Score on D2: 0.08385670937519984 | D1-D2 diff: 1.8128062338780866 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), UnderDominanceEncoder(), RecessiveEncoder(), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=20, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17043859344528645 Holdout data R^2 trained on entire dataset(80%): 0.09980762851710001 Dataset D1 R^2 on trained D1: 0.17645318480291416 .................................................. Pipeline #4: Score on D2: 0.0765797200704641 | D1-D2 diff: 1.8381799633953175 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1655805684223659 Holdout data R^2 trained on entire dataset(80%): 0.09679428031531678 Dataset D1 R^2 on trained D1: 0.16416838133465605 .................................................. Pipeline #5: Score on D2: 0.06844345607733016 | D1-D2 diff: 1.867507348601446 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1638509606665628 Holdout data R^2 trained on entire dataset(80%): 0.09207027068796358 Dataset D1 R^2 on trained D1: 0.15065839193084862 .................................................. Pipeline #6: Score on D2: 0.061967294802885986 | D1-D2 diff: 1.8919850854405493 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1412044668720176 Holdout data R^2 trained on entire dataset(80%): 0.08109693087323422 Dataset D1 R^2 on trained D1: 0.14000943411704703 .................................................. Pipeline #7: Score on D2: 0.05950879364637074 | D1-D2 diff: 1.923818636374813 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RecessiveEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14120194176749168 Holdout data R^2 trained on entire dataset(80%): 0.08325163066938202 Dataset D1 R^2 on trained D1: 0.13251226137370753 .................................................. Pipeline #8: Score on D2: 0.059332113140699194 | D1-D2 diff: 1.9784936722934894 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1326095450735525 Holdout data R^2 trained on entire dataset(80%): 0.08072987764709572 Dataset D1 R^2 on trained D1: 0.12459425718887751 .................................................. Pipeline #9: Score on D2: 0.04984659177119555 | D1-D2 diff: 2.015960461487535 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.2), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11800002981150137 Holdout data R^2 trained on entire dataset(80%): 0.06608967680856426 Dataset D1 R^2 on trained D1: 0.11039071012968671 .................................................. Pipeline #10: Score on D2: 0.04839343436328292 | D1-D2 diff: 2.028211252966575 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1066356685679859 Holdout data R^2 trained on entire dataset(80%): 0.05693105292523082 Dataset D1 R^2 on trained D1: 0.10748795964233615 .................................................. Pipeline #11: Score on D2: 0.04661011019634753 | D1-D2 diff: 2.099959262876287 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), DominantEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10214210192892137 Holdout data R^2 trained on entire dataset(80%): 0.05705737977763359 Dataset D1 R^2 on trained D1: 0.09803300488963751 .................................................. Pipeline #12: Score on D2: 0.044633765257121594 | D1-D2 diff: 2.120497782583612 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10810786417043927 Holdout data R^2 trained on entire dataset(80%): 0.060832260557059925 Dataset D1 R^2 on trained D1: 0.09409314997751062 .................................................. Pipeline #13: Score on D2: 0.03568990743063938 | D1-D2 diff: 2.2193387176880477 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08808142542901765 Holdout data R^2 trained on entire dataset(80%): 0.04834705957972607 Dataset D1 R^2 on trained D1: 0.07690968472528692 .................................................. Pipeline #14: Score on D2: 0.03535840050194372 | D1-D2 diff: 2.279883306278217 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08376060146420294 Holdout data R^2 trained on entire dataset(80%): 0.047373948122774845 Dataset D1 R^2 on trained D1: 0.07237099468984431 .................................................. Pipeline #15: Score on D2: 0.007904705112022126 | D1-D2 diff: 2.3946630328807053 Pipeline steps: DominantEncoder(), VarianceThreshold(threshold=0.05), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.042210967590551096 Holdout data R^2 trained on entire dataset(80%): 0.013579752839382286 Dataset D1 R^2 on trained D1: 0.038315121055812584 .................................................. Pipeline #16: Score on D2: 0.00783294463869344 | D1-D2 diff: 2.524037769986081 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=90), HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.018979143061331527 Holdout data R^2 trained on entire dataset(80%): 0.0018421692849766025 Dataset D1 R^2 on trained D1: 0.03247157727248273 .................................................. Pipeline #17: Score on D2: 0.004875245765563085 | D1-D2 diff: 2.6225541440142366 Pipeline steps: UnderDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.026830079613965552 Holdout data R^2 trained on entire dataset(80%): 0.009792757652894224 Dataset D1 R^2 on trained D1: 0.026015107730766096 .................................................. Pipeline #18: Score on D2: 0.004402608062586744 | D1-D2 diff: 2.7114067461922144 Pipeline steps: UnderDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.02480195483483283 Holdout data R^2 trained on entire dataset(80%): 0.00814249973712311 Dataset D1 R^2 on trained D1: 0.022904720289674607 .................................................. Pipeline #19: Score on D2: 9.980889340099885e-05 | D1-D2 diff: 5.573115327046485 Pipeline steps: RecessiveEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), DecisionTreeRegressor(max_depth=1, min_samples_leaf=7, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0007692882247953836 Holdout data R^2 trained on entire dataset(80%): -0.0015462237301027137 Dataset D1 R^2 on trained D1: 0.0011364009563482602 .................................................. Pipeline #20: Score on D2: 1.9373274260048312e-05 | D1-D2 diff: 14.085515569901522 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), DominantEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08808142542901765 Holdout data R^2 trained on entire dataset(80%): 0.04834705957972607 Dataset D1 R^2 on trained D1: -6.031129581973715e-06 .................................................. ************************************************************************************** Random Seed 31 - 3 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.12625 Best score on D1-D2 diff: 7.01058 Gen 2 - Best score on D2: 0.12987 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.13063 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.13063 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.13063 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.13063 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.13063 Best score on D1-D2 diff: 16.36782 Gen 8 - Best score on D2: 0.13063 Best score on D1-D2 diff: 16.36782 Gen 9 - Best score on D2: 0.13063 Best score on D1-D2 diff: 16.36782 Gen 10 - Best score on D2: 0.13063 Best score on D1-D2 diff: 16.36782 Gen 11 - Best score on D2: 0.13097 Best score on D1-D2 diff: 16.36782 Gen 12 - Best score on D2: 0.13097 Best score on D1-D2 diff: 16.36782 Gen 13 - Best score on D2: 0.13097 Best score on D1-D2 diff: 16.36782 Gen 14 - Best score on D2: 0.13097 Best score on D1-D2 diff: 16.36782 Gen 15 - Best score on D2: 0.13097 Best score on D1-D2 diff: 16.36782 Gen 16 - Best score on D2: 0.13097 Best score on D1-D2 diff: 16.36782 Gen 17 - Best score on D2: 0.13097 Best score on D1-D2 diff: 16.36782 Gen 18 - Best score on D2: 0.13097 Best score on D1-D2 diff: 16.36782 Gen 19 - Best score on D2: 0.13097 Best score on D1-D2 diff: 16.36782 Gen 20 - Best score on D2: 0.13225 Best score on D1-D2 diff: 16.36782 Gen 21 - Best score on D2: 0.13225 Best score on D1-D2 diff: 16.36782 Gen 22 - Best score on D2: 0.13225 Best score on D1-D2 diff: 16.36782 Gen 23 - Best score on D2: 0.13225 Best score on D1-D2 diff: 16.36782 Gen 24 - Best score on D2: 0.13225 Best score on D1-D2 diff: 16.36782 Gen 25 - Best score on D2: 0.13225 Best score on D1-D2 diff: 16.36782 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.008748145241207261 Entire dataset(80%) R^2 trained on data (80%): 0.002811579136402509 Holdout R^2 (20%) trained on data (80%): -0.008162616959733526 Dataset D1 R^2 on trained D1: 0.006279588921648149 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0019318502209823007 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1322529229217151 | D1-D2 diff: 1.462292003353768 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=8, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3311492782926302 Holdout data R^2 trained on entire dataset(80%): 0.15471627368389906 Dataset D1 R^2 on trained D1: 0.35096037437131444 .................................................. Pipeline #2: Score on D2: 0.1309652185387814 | D1-D2 diff: 1.6740661659474085 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24628061647355948 Holdout data R^2 trained on entire dataset(80%): 0.1555111000025089 Dataset D1 R^2 on trained D1: 0.2582889985730802 .................................................. Pipeline #3: Score on D2: 0.12957237775489383 | D1-D2 diff: 1.6882298886022438 Pipeline steps: VarianceThreshold(threshold=0.25), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2433223358572454 Holdout data R^2 trained on entire dataset(80%): 0.14824312983086352 Dataset D1 R^2 on trained D1: 0.2526768018597696 .................................................. Pipeline #4: Score on D2: 0.12956608883264187 | D1-D2 diff: 1.7160370027020913 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=19, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23722300891355763 Holdout data R^2 trained on entire dataset(80%): 0.15432684509560324 Dataset D1 R^2 on trained D1: 0.24488310779081712 .................................................. Pipeline #5: Score on D2: 0.127743270005236 | D1-D2 diff: 1.7245571960358284 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=20, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23325302134022696 Holdout data R^2 trained on entire dataset(80%): 0.15562094641413438 Dataset D1 R^2 on trained D1: 0.2407982219225604 .................................................. Pipeline #6: Score on D2: 0.12753249619017426 | D1-D2 diff: 1.7256027932496212 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=20, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23194333882663132 Holdout data R^2 trained on entire dataset(80%): 0.154975804604916 Dataset D1 R^2 on trained D1: 0.24031368279862897 .................................................. Pipeline #7: Score on D2: 0.12675825646594618 | D1-D2 diff: 1.732538686705171 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23268040764369602 Holdout data R^2 trained on entire dataset(80%): 0.15087632451553012 Dataset D1 R^2 on trained D1: 0.2377442658404718 .................................................. Pipeline #8: Score on D2: 0.12635616536830996 | D1-D2 diff: 1.8210187890081908 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.4, min_samples_leaf=20, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21099905225128113 Holdout data R^2 trained on entire dataset(80%): 0.1464469867901229 Dataset D1 R^2 on trained D1: 0.21729351504881067 .................................................. Pipeline #9: Score on D2: 0.12024434368330394 | D1-D2 diff: 1.8456004751821253 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20910212604246436 Holdout data R^2 trained on entire dataset(80%): 0.14205548538944324 Dataset D1 R^2 on trained D1: 0.2064328246792254 .................................................. Pipeline #10: Score on D2: 0.11892693251895237 | D1-D2 diff: 1.9400155031106392 Pipeline steps: VarianceThreshold(threshold=0.25), UnderDominanceEncoder(), RecessiveEncoder(), SelectPercentile(percentile=50), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13639549816955654 Holdout data R^2 trained on entire dataset(80%): 0.08545263878815912 Dataset D1 R^2 on trained D1: 0.1895227868062772 .................................................. Pipeline #11: Score on D2: 0.11371962508396072 | D1-D2 diff: 1.97338470265078 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), SelectPercentile(percentile=50), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12787756426088703 Holdout data R^2 trained on entire dataset(80%): 0.08613830906872222 Dataset D1 R^2 on trained D1: 0.1796602366663037 .................................................. Pipeline #12: Score on D2: 0.10784635382272201 | D1-D2 diff: 2.026110806309444 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), SelectPercentile(percentile=50), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12082167047767056 Holdout data R^2 trained on entire dataset(80%): 0.08543983294584767 Dataset D1 R^2 on trained D1: 0.16718631098928594 .................................................. Pipeline #13: Score on D2: 0.10676942341933227 | D1-D2 diff: 2.1022323677687598 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), SelectPercentile(percentile=50), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11820778728714654 Holdout data R^2 trained on entire dataset(80%): 0.08482606704948004 Dataset D1 R^2 on trained D1: 0.1579702680937234 .................................................. Pipeline #14: Score on D2: 0.09164493492121994 | D1-D2 diff: 2.2384769110431875 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=3, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11274544387922514 Holdout data R^2 trained on entire dataset(80%): 0.12049855794747866 Dataset D1 R^2 on trained D1: 0.13147302890671364 .................................................. Pipeline #15: Score on D2: 0.06003322587792426 | D1-D2 diff: 2.2716851294663867 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11402764453089387 Holdout data R^2 trained on entire dataset(80%): 0.07959517552146222 Dataset D1 R^2 on trained D1: 0.09758301129094626 .................................................. Pipeline #16: Score on D2: 0.0448349829002459 | D1-D2 diff: 2.7963476663300084 Pipeline steps: UnderDominanceEncoder(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.057873832305418316 Holdout data R^2 trained on entire dataset(80%): 0.061261601143761 Dataset D1 R^2 on trained D1: 0.06118941007152923 .................................................. Pipeline #17: Score on D2: 0.024175634237669885 | D1-D2 diff: 2.9040485132894385 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=75), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.02836594538828474 Holdout data R^2 trained on entire dataset(80%): 0.007002356799045706 Dataset D1 R^2 on trained D1: 0.03823560868497666 .................................................. Pipeline #18: Score on D2: 0.023644591714986674 | D1-D2 diff: 2.950332903072396 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=75), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.029971915417032147 Holdout data R^2 trained on entire dataset(80%): 0.008496072621744344 Dataset D1 R^2 on trained D1: 0.03684282830557939 .................................................. Pipeline #19: Score on D2: 0.015891168141994316 | D1-D2 diff: 3.0111668872849626 Pipeline steps: DominantEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03983088312675431 Holdout data R^2 trained on entire dataset(80%): 0.020702242037343832 Dataset D1 R^2 on trained D1: 0.028054727981571492 .................................................. Pipeline #20: Score on D2: 0.005109467629911824 | D1-D2 diff: 5.306857007837277 Pipeline steps: SelectPercentile(percentile=35), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=16, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0023054372444346605 Holdout data R^2 trained on entire dataset(80%): -0.002152460426459557 Dataset D1 R^2 on trained D1: 0.0063702799873215366 .................................................. Pipeline #21: Score on D2: 0.004029551413889498 | D1-D2 diff: 7.010575836964466 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004258656997543908 Holdout data R^2 trained on entire dataset(80%): 0.005549324780568465 Dataset D1 R^2 on trained D1: 0.0044435370123858675 .................................................. Pipeline #22: Score on D2: 0.0040292704173320315 | D1-D2 diff: 7.018173054535042 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=5), RecessiveEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004256908935060033 Holdout data R^2 trained on entire dataset(80%): 0.005445908017993606 Dataset D1 R^2 on trained D1: 0.0044414663561128664 .................................................. Pipeline #23: Score on D2: 0.0010110497422896048 | D1-D2 diff: 16.367823692588143 Pipeline steps: SelectPercentile(percentile=20), HeterosisEncoder(), SelectPercentile(percentile=25), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=15, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0006899776717492756 Holdout data R^2 trained on entire dataset(80%): -0.002429208158663876 Dataset D1 R^2 on trained D1: 0.0009971170114710937 .................................................. Pipeline #24: Score on D2: 0.0010110497422893827 | D1-D2 diff: 16.36782369271857 Pipeline steps: SelectPercentile(percentile=20), HeterosisEncoder(), RecessiveEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0006899776717492756 Holdout data R^2 trained on entire dataset(80%): -0.002429208158663876 Dataset D1 R^2 on trained D1: 0.0009971170114713157 .................................................. ************************************************************************************** Random Seed 31 - 4 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.12823 Best score on D1-D2 diff: 7.01058 Gen 2 - Best score on D2: 0.12958 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.13373 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.13523 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.13523 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.13523 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.13530 Best score on D1-D2 diff: 20.63568 Gen 8 - Best score on D2: 0.13530 Best score on D1-D2 diff: 20.63568 Gen 9 - Best score on D2: 0.13530 Best score on D1-D2 diff: 20.63568 Gen 10 - Best score on D2: 0.13530 Best score on D1-D2 diff: 20.63568 Gen 11 - Best score on D2: 0.13530 Best score on D1-D2 diff: 20.63568 Gen 12 - Best score on D2: 0.13717 Best score on D1-D2 diff: 20.63568 Gen 13 - Best score on D2: 0.13717 Best score on D1-D2 diff: 20.63568 Gen 14 - Best score on D2: 0.13717 Best score on D1-D2 diff: 20.63568 Gen 15 - Best score on D2: 0.13717 Best score on D1-D2 diff: 20.63568 Gen 16 - Best score on D2: 0.13854 Best score on D1-D2 diff: 20.63568 Gen 17 - Best score on D2: 0.13854 Best score on D1-D2 diff: 20.63568 Gen 18 - Best score on D2: 0.13854 Best score on D1-D2 diff: 20.63568 Gen 19 - Best score on D2: 0.13854 Best score on D1-D2 diff: 20.63568 Gen 20 - Best score on D2: 0.13854 Best score on D1-D2 diff: 20.63568 Gen 21 - Best score on D2: 0.13854 Best score on D1-D2 diff: 20.63568 Gen 22 - Best score on D2: 0.13854 Best score on D1-D2 diff: 20.63568 Gen 23 - Best score on D2: 0.13854 Best score on D1-D2 diff: 20.63568 Gen 24 - Best score on D2: 0.13854 Best score on D1-D2 diff: 20.63568 Gen 25 - Best score on D2: 0.13854 Best score on D1-D2 diff: 20.63568 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.006196977654475466 Entire dataset(80%) R^2 trained on data (80%): 0.0028865835668253625 Holdout R^2 (20%) trained on data (80%): -0.0051143806929694335 Dataset D1 R^2 on trained D1: 0.005753664540789405 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0023085144288562676 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.13853676633505407 | D1-D2 diff: 1.432333968138217 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=4, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.36649497847779233 Holdout data R^2 trained on entire dataset(80%): 0.17207868327556708 Dataset D1 R^2 on trained D1: 0.3761238541261458 .................................................. Pipeline #2: Score on D2: 0.13756834266970297 | D1-D2 diff: 1.785223481872426 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=95), RandomForestRegressor(max_features=0.55, min_samples_leaf=20, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23372757310753567 Holdout data R^2 trained on entire dataset(80%): 0.169790208943562 Dataset D1 R^2 on trained D1: 0.23602149751150459 .................................................. Pipeline #3: Score on D2: 0.13154910116027863 | D1-D2 diff: 1.7876929853813972 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22713784973283824 Holdout data R^2 trained on entire dataset(80%): 0.16370926441581424 Dataset D1 R^2 on trained D1: 0.2294593728789046 .................................................. Pipeline #4: Score on D2: 0.13073078233444935 | D1-D2 diff: 1.7907685566332607 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22971856795295176 Holdout data R^2 trained on entire dataset(80%): 0.16179542805710911 Dataset D1 R^2 on trained D1: 0.2279701574459324 .................................................. Pipeline #5: Score on D2: 0.12902406406057876 | D1-D2 diff: 1.851760413400995 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=20, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21634586622038987 Holdout data R^2 trained on entire dataset(80%): 0.1611265036526609 Dataset D1 R^2 on trained D1: 0.2140714202692765 .................................................. Pipeline #6: Score on D2: 0.11306859968181815 | D1-D2 diff: 2.03422979056608 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17499872813152018 Holdout data R^2 trained on entire dataset(80%): 0.13881653856336307 Dataset D1 R^2 on trained D1: 0.17146686670595335 .................................................. Pipeline #7: Score on D2: 0.09164493492121994 | D1-D2 diff: 2.2384769110431875 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=75), DecisionTreeRegressor(max_depth=2, min_samples_leaf=11, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11274544387922503 Holdout data R^2 trained on entire dataset(80%): 0.12049855794747866 Dataset D1 R^2 on trained D1: 0.13147302890671364 .................................................. Pipeline #8: Score on D2: 0.07087768774943248 | D1-D2 diff: 2.2730550698548777 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=95), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1259493728170914 Holdout data R^2 trained on entire dataset(80%): 0.09669162279196042 Dataset D1 R^2 on trained D1: 0.10833703191835353 .................................................. Pipeline #9: Score on D2: 0.05381243287420745 | D1-D2 diff: 2.355229100581206 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=35), VarianceThreshold(threshold=0.1), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=5, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.010646947726984823 Holdout data R^2 trained on entire dataset(80%): 0.0014136160633504424 Dataset D1 R^2 on trained D1: 0.08631123586277267 .................................................. Pipeline #10: Score on D2: 0.045651392286603754 | D1-D2 diff: 2.7703761914868332 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=35), FeatureEncodingFrequencySelector(threshold=0.0), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=8, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.008809304234861925 Holdout data R^2 trained on entire dataset(80%): 0.003588213797646511 Dataset D1 R^2 on trained D1: 0.06262776943630399 .................................................. Pipeline #11: Score on D2: 0.01808536594405541 | D1-D2 diff: 2.854975937297427 Pipeline steps: DominantEncoder(), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03876607513908392 Holdout data R^2 trained on entire dataset(80%): 0.021208092069863405 Dataset D1 R^2 on trained D1: 0.03313722666898822 .................................................. Pipeline #12: Score on D2: 0.004469256237887076 | D1-D2 diff: 2.875492355727678 Pipeline steps: DominantEncoder(), HeterosisEncoder(), HeterosisEncoder(), SelectPercentile(percentile=70), RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0514288221387581 Holdout data R^2 trained on entire dataset(80%): 0.022960800905027057 Dataset D1 R^2 on trained D1: 0.019096117107469945 .................................................. Pipeline #13: Score on D2: 0.004029551413889498 | D1-D2 diff: 7.010575836964466 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=35), DecisionTreeRegressor(max_depth=1, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004258656997543686 Holdout data R^2 trained on entire dataset(80%): 0.005549324780568465 Dataset D1 R^2 on trained D1: 0.0044435370123858675 .................................................. Pipeline #14: Score on D2: 0.0040292704173320315 | D1-D2 diff: 7.018173054535042 Pipeline steps: VarianceThreshold(), HeterosisEncoder(), SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.05, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.004256908935060033 Holdout data R^2 trained on entire dataset(80%): 0.005445908017993606 Dataset D1 R^2 on trained D1: 0.0044414663561128664 .................................................. Pipeline #15: Score on D2: 0.0010110497422893827 | D1-D2 diff: 16.36782369271857 Pipeline steps: SelectPercentile(percentile=20), HeterosisEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.001026270207570068 Holdout data R^2 trained on entire dataset(80%): -0.0006436442563220179 Dataset D1 R^2 on trained D1: 0.0009971170114713157 .................................................. Pipeline #16: Score on D2: -6.674184258592675e-07 | D1-D2 diff: 20.635678995370764 Pipeline steps: SelectPercentile(percentile=5), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0002597503345875829 Holdout data R^2 trained on entire dataset(80%): -0.0016525566401370817 Dataset D1 R^2 on trained D1: 4.847320359435692e-06 .................................................. ************************************************************************************** Random Seed 31 - 5 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.14283 Best score on D1-D2 diff: 3.61179 Gen 2 - Best score on D2: 0.14830 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14830 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.14981 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.15147 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.013150537604913248 Entire dataset(80%) R^2 trained on data (80%): 0.002799660453520092 Holdout R^2 (20%) trained on data (80%): -0.007957315398312126 Dataset D1 R^2 on trained D1: 0.008481370206462335 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0015639163981847615 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.15147006009419917 | D1-D2 diff: 1.3967429678510412 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=2, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4081614488005906 Holdout data R^2 trained on entire dataset(80%): 0.19699391769492125 Dataset D1 R^2 on trained D1: 0.4142147973831113 .................................................. Pipeline #2: Score on D2: 0.15116617206646055 | D1-D2 diff: 1.4945713984085802 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=8, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34252671798615497 Holdout data R^2 trained on entire dataset(80%): 0.1901759421229613 Dataset D1 R^2 on trained D1: 0.35158260692819154 .................................................. Pipeline #3: Score on D2: 0.14798177357050535 | D1-D2 diff: 1.5656601278349174 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=8, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30913432222342874 Holdout data R^2 trained on entire dataset(80%): 0.18774309529720168 Dataset D1 R^2 on trained D1: 0.3144035039465015 .................................................. Pipeline #4: Score on D2: 0.14545816615842677 | D1-D2 diff: 1.7303563536735105 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2533851124564884 Holdout data R^2 trained on entire dataset(80%): 0.17832385492473113 Dataset D1 R^2 on trained D1: 0.25700513982379003 .................................................. Pipeline #5: Score on D2: 0.1422860965896977 | D1-D2 diff: 1.7746882562254052 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24096643017876063 Holdout data R^2 trained on entire dataset(80%): 0.16937992996913753 Dataset D1 R^2 on trained D1: 0.24309797299258007 .................................................. Pipeline #6: Score on D2: 0.1400713711842968 | D1-D2 diff: 1.7842852746865379 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23850931720110158 Holdout data R^2 trained on entire dataset(80%): 0.16786594113987008 Dataset D1 R^2 on trained D1: 0.2387317627013522 .................................................. Pipeline #7: Score on D2: 0.13931038822022257 | D1-D2 diff: 1.8188343944962049 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22620403360499752 Holdout data R^2 trained on entire dataset(80%): 0.1661788513384187 Dataset D1 R^2 on trained D1: 0.23068538343027367 .................................................. Pipeline #8: Score on D2: 0.1298680302429609 | D1-D2 diff: 1.8319315746510907 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.2), FeatureEncodingFrequencySelector(threshold=0.3), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21885892287984565 Holdout data R^2 trained on entire dataset(80%): 0.16781124678658277 Dataset D1 R^2 on trained D1: 0.218657815908191 .................................................. Pipeline #9: Score on D2: 0.12436769709773698 | D1-D2 diff: 1.8545596127333939 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21307363595277573 Holdout data R^2 trained on entire dataset(80%): 0.16341266564420742 Dataset D1 R^2 on trained D1: 0.2089027461140579 .................................................. Pipeline #10: Score on D2: 0.11320073568878097 | D1-D2 diff: 1.970124120797347 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18090058761279237 Holdout data R^2 trained on entire dataset(80%): 0.13283481627342353 Dataset D1 R^2 on trained D1: 0.1795789625504023 .................................................. Pipeline #11: Score on D2: 0.11316820196072597 | D1-D2 diff: 1.9855415455282117 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1823504060679847 Holdout data R^2 trained on entire dataset(80%): 0.13667130036543962 Dataset D1 R^2 on trained D1: 0.17750865065814492 .................................................. Pipeline #12: Score on D2: 0.09623664233961671 | D1-D2 diff: 2.1080412760445855 Pipeline steps: SelectPercentile(percentile=25), HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.118740697285585 Holdout data R^2 trained on entire dataset(80%): 0.11963868355431295 Dataset D1 R^2 on trained D1: 0.14687546024015896 .................................................. Pipeline #13: Score on D2: 0.09306082686969686 | D1-D2 diff: 2.366604444228707 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10983575359514564 Holdout data R^2 trained on entire dataset(80%): 0.07916314262171287 Dataset D1 R^2 on trained D1: 0.12493928422138334 .................................................. Pipeline #14: Score on D2: 0.04807135589802358 | D1-D2 diff: 2.768089464414847 Pipeline steps: VarianceThreshold(threshold=0.15), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.45, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06365941643152173 Holdout data R^2 trained on entire dataset(80%): 0.052983936303906964 Dataset D1 R^2 on trained D1: 0.06510389954467011 .................................................. Pipeline #15: Score on D2: 0.03837007568299311 | D1-D2 diff: 2.844916564538438 Pipeline steps: SelectPercentile(percentile=25), HeterosisEncoder(), HeterosisEncoder(), SelectPercentile(percentile=20), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.003528907950673621 Holdout data R^2 trained on entire dataset(80%): 0.002042333437447308 Dataset D1 R^2 on trained D1: 0.05363595638301144 .................................................. Pipeline #16: Score on D2: 0.038226370554786615 | D1-D2 diff: 5.81081420206724 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=5), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03779870685399844 Holdout data R^2 trained on entire dataset(80%): 0.015662869637106724 Dataset D1 R^2 on trained D1: 0.03734926463379118 .................................................. Pipeline #17: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 31 - 6 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.15045 Best score on D1-D2 diff: 4.81299 Gen 2 - Best score on D2: 0.15045 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.15272 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15272 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.15445 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.15445 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15604 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.15604 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.15604 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.15773 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.15775 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.15837 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.15837 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.15837 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.15892 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.15892 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.15892 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.16161 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.16161 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.16161 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.16161 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.16199 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.16199 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.16199 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.16199 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.011942697610668462 Entire dataset(80%) R^2 trained on data (80%): 0.0028047847561785133 Holdout R^2 (20%) trained on data (80%): -0.005412817184780261 Dataset D1 R^2 on trained D1: 0.007273154384504488 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0019397473448315994 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.16199016117263565 | D1-D2 diff: 1.5168063047117306 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=3, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34499182034685694 Holdout data R^2 trained on entire dataset(80%): 0.20818565239364262 Dataset D1 R^2 on trained D1: 0.35091084082579893 .................................................. Pipeline #2: Score on D2: 0.15741019601366957 | D1-D2 diff: 1.522797216227645 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), RandomForestRegressor(max_features=0.3, min_samples_leaf=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33600017934819426 Holdout data R^2 trained on entire dataset(80%): 0.1983422019875527 Dataset D1 R^2 on trained D1: 0.3433754052555028 .................................................. Pipeline #3: Score on D2: 0.15508158682251827 | D1-D2 diff: 1.5967024440793338 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), DominantEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=3, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3089913009168458 Holdout data R^2 trained on entire dataset(80%): 0.193185089318473 Dataset D1 R^2 on trained D1: 0.3089339033498746 .................................................. Pipeline #4: Score on D2: 0.15336324671597423 | D1-D2 diff: 1.6815913277399757 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=20, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2696402108664193 Holdout data R^2 trained on entire dataset(80%): 0.1875014953567139 Dataset D1 R^2 on trained D1: 0.278423171795784 .................................................. Pipeline #5: Score on D2: 0.14840773072987745 | D1-D2 diff: 1.6871424761598695 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2668126797012442 Holdout data R^2 trained on entire dataset(80%): 0.18816607082147752 Dataset D1 R^2 on trained D1: 0.27182983934593796 .................................................. Pipeline #6: Score on D2: 0.14820272144835311 | D1-D2 diff: 1.703030697585263 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), RandomForestRegressor(max_features=0.5, min_samples_leaf=19, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2614481662662549 Holdout data R^2 trained on entire dataset(80%): 0.18791043190253653 Dataset D1 R^2 on trained D1: 0.26708307684865895 .................................................. Pipeline #7: Score on D2: 0.14750636841152054 | D1-D2 diff: 1.778331440098648 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24806524358423565 Holdout data R^2 trained on entire dataset(80%): 0.18334044903039504 Dataset D1 R^2 on trained D1: 0.24749466580644497 .................................................. Pipeline #8: Score on D2: 0.145688656037116 | D1-D2 diff: 1.8149131294740797 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23184500523683627 Holdout data R^2 trained on entire dataset(80%): 0.17733697933909032 Dataset D1 R^2 on trained D1: 0.23785590617567698 .................................................. Pipeline #9: Score on D2: 0.13809569163776847 | D1-D2 diff: 1.859127100282092 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=16, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22156289201002377 Holdout data R^2 trained on entire dataset(80%): 0.16636253765311937 Dataset D1 R^2 on trained D1: 0.22180305713539794 .................................................. Pipeline #10: Score on D2: 0.13795083009139075 | D1-D2 diff: 1.8720821310970983 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21721389151846393 Holdout data R^2 trained on entire dataset(80%): 0.16768698887681865 Dataset D1 R^2 on trained D1: 0.21936507664763893 .................................................. Pipeline #11: Score on D2: 0.13780338217297738 | D1-D2 diff: 1.8811648322874217 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.15), OverDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=19, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2112075044109707 Holdout data R^2 trained on entire dataset(80%): 0.16519994780038283 Dataset D1 R^2 on trained D1: 0.21765663196285334 .................................................. Pipeline #12: Score on D2: 0.1373070928953759 | D1-D2 diff: 1.900586585334377 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.05), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.2, min_samples_leaf=19, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21369010797631838 Holdout data R^2 trained on entire dataset(80%): 0.1624625143827909 Dataset D1 R^2 on trained D1: 0.2139460103419184 .................................................. Pipeline #13: Score on D2: 0.13397052738063153 | D1-D2 diff: 1.9039240879030348 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21195635056703754 Holdout data R^2 trained on entire dataset(80%): 0.16263275118132425 Dataset D1 R^2 on trained D1: 0.21007347638981289 .................................................. Pipeline #14: Score on D2: 0.13210576033559507 | D1-D2 diff: 1.9293780441761288 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20451792825338055 Holdout data R^2 trained on entire dataset(80%): 0.16067874120750414 Dataset D1 R^2 on trained D1: 0.2042714342770926 .................................................. Pipeline #15: Score on D2: 0.12249516515528791 | D1-D2 diff: 1.9846360224006434 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18903197018767148 Holdout data R^2 trained on entire dataset(80%): 0.14641914762273012 Dataset D1 R^2 on trained D1: 0.18695311983395768 .................................................. Pipeline #16: Score on D2: 0.12005541334128111 | D1-D2 diff: 2.001411496818882 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18768290726133552 Holdout data R^2 trained on entire dataset(80%): 0.14843499721221398 Dataset D1 R^2 on trained D1: 0.18237928710057416 .................................................. Pipeline #17: Score on D2: 0.11607038545490511 | D1-D2 diff: 2.1003141616576486 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=12, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15532581077523688 Holdout data R^2 trained on entire dataset(80%): 0.131903362068166 Dataset D1 R^2 on trained D1: 0.16745853239880149 .................................................. Pipeline #18: Score on D2: 0.1154091726115617 | D1-D2 diff: 2.6881132497629 Pipeline steps: SelectPercentile(percentile=20), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=20, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.112911933249501 Holdout data R^2 trained on entire dataset(80%): 0.11942608562325574 Dataset D1 R^2 on trained D1: 0.13456097949602586 .................................................. Pipeline #19: Score on D2: 0.048164332483421 | D1-D2 diff: 2.7717275456604447 Pipeline steps: DominantEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06365941643152173 Holdout data R^2 trained on entire dataset(80%): 0.052983936303906964 Dataset D1 R^2 on trained D1: 0.06510762654568192 .................................................. Pipeline #20: Score on D2: 0.044643732943826286 | D1-D2 diff: 6.600018874056713 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=50), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04440095216163509 Holdout data R^2 trained on entire dataset(80%): 0.029562242855793897 Dataset D1 R^2 on trained D1: 0.04411672241716624 .................................................. Pipeline #21: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=8, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 31 - 7 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.16997 Best score on D1-D2 diff: 4.43369 Gen 2 - Best score on D2: 0.16997 Best score on D1-D2 diff: 5.43283 Gen 3 - Best score on D2: 0.17509 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.17509 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.17509 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.17509 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.17678 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.17678 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.17811 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.18226 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.18226 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.18253 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.18253 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.18253 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.18253 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.18253 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.00897210900340828 Entire dataset(80%) R^2 trained on data (80%): 0.002196955119277977 Holdout R^2 (20%) trained on data (80%): 0.0027659673271511753 Dataset D1 R^2 on trained D1: 0.004377251526767267 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0032417287800479144 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1825333593061359 | D1-D2 diff: 1.5544910867718644 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=3, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3501060629293957 Holdout data R^2 trained on entire dataset(80%): 0.2181560797484856 Dataset D1 R^2 on trained D1: 0.35378985582064404 .................................................. Pipeline #2: Score on D2: 0.1745189644149916 | D1-D2 diff: 1.5759775529584725 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=8, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.32905495273763574 Holdout data R^2 trained on entire dataset(80%): 0.21005575767367457 Dataset D1 R^2 on trained D1: 0.33662526327167486 .................................................. Pipeline #3: Score on D2: 0.17407490505885115 | D1-D2 diff: 1.5822387124436619 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.32959338676018723 Holdout data R^2 trained on entire dataset(80%): 0.21277945265070208 Dataset D1 R^2 on trained D1: 0.33363047714036176 .................................................. Pipeline #4: Score on D2: 0.17327050886350148 | D1-D2 diff: 1.593242981044059 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=8, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3244590408340058 Holdout data R^2 trained on entire dataset(80%): 0.20791669508101818 Dataset D1 R^2 on trained D1: 0.3284634431026947 .................................................. Pipeline #5: Score on D2: 0.17265915208667704 | D1-D2 diff: 1.6236287306554442 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=17, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3103773987570656 Holdout data R^2 trained on entire dataset(80%): 0.20793598429275617 Dataset D1 R^2 on trained D1: 0.31655659920666646 .................................................. Pipeline #6: Score on D2: 0.1708889506127511 | D1-D2 diff: 1.6430013364955487 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=12, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30645928418716295 Holdout data R^2 trained on entire dataset(80%): 0.20701933968260844 Dataset D1 R^2 on trained D1: 0.3081187198865287 .................................................. Pipeline #7: Score on D2: 0.16997466713225273 | D1-D2 diff: 1.6527712205436786 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29903124214554655 Holdout data R^2 trained on entire dataset(80%): 0.2054180658959064 Dataset D1 R^2 on trained D1: 0.3039883160341046 .................................................. Pipeline #8: Score on D2: 0.16950222190657038 | D1-D2 diff: 1.6676556557938416 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=8, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2927327899433221 Holdout data R^2 trained on entire dataset(80%): 0.19774814171801414 Dataset D1 R^2 on trained D1: 0.2987950625295347 .................................................. Pipeline #9: Score on D2: 0.1687491388193032 | D1-D2 diff: 1.6807235203196955 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=19, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2879297532396551 Holdout data R^2 trained on entire dataset(80%): 0.20340126601607944 Dataset D1 R^2 on trained D1: 0.29406755261037065 .................................................. Pipeline #10: Score on D2: 0.16833324482946466 | D1-D2 diff: 1.6870440421473403 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=19, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2885920826704491 Holdout data R^2 trained on entire dataset(80%): 0.2017346809733962 Dataset D1 R^2 on trained D1: 0.29178416122150497 .................................................. Pipeline #11: Score on D2: 0.1659282600521773 | D1-D2 diff: 1.7011307137763612 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2833195241380926 Holdout data R^2 trained on entire dataset(80%): 0.20088374144284193 Dataset D1 R^2 on trained D1: 0.2853406132377658 .................................................. Pipeline #12: Score on D2: 0.16447095914857268 | D1-D2 diff: 1.73373305118669 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=17, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2761449174210504 Holdout data R^2 trained on entire dataset(80%): 0.19968199927504582 Dataset D1 R^2 on trained D1: 0.27515145244082606 .................................................. Pipeline #13: Score on D2: 0.16332292926909664 | D1-D2 diff: 1.7778330998557446 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2596643012234561 Holdout data R^2 trained on entire dataset(80%): 0.19469075422297732 Dataset D1 R^2 on trained D1: 0.26342338375495844 .................................................. Pipeline #14: Score on D2: 0.16282329218950597 | D1-D2 diff: 1.7886217936968867 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=20, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26261374607555465 Holdout data R^2 trained on entire dataset(80%): 0.19741700262677953 Dataset D1 R^2 on trained D1: 0.2605303480834755 .................................................. Pipeline #15: Score on D2: 0.16223545515586335 | D1-D2 diff: 1.8301160936874892 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24668017207350323 Holdout data R^2 trained on entire dataset(80%): 0.1857572468507228 Dataset D1 R^2 on trained D1: 0.2513780844150547 .................................................. Pipeline #16: Score on D2: 0.15885729569333285 | D1-D2 diff: 1.8669392943604242 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23809273804176356 Holdout data R^2 trained on entire dataset(80%): 0.18239141948393112 Dataset D1 R^2 on trained D1: 0.2411723394887637 .................................................. Pipeline #17: Score on D2: 0.13815463110458037 | D1-D2 diff: 1.8818143556674702 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RandomForestRegressor(bootstrap=False, max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2209993105041106 Holdout data R^2 trained on entire dataset(80%): 0.16302790755628727 Dataset D1 R^2 on trained D1: 0.21789768999173265 .................................................. Pipeline #18: Score on D2: 0.13130211036241668 | D1-D2 diff: 1.8991906451543177 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=10, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21119405916836786 Holdout data R^2 trained on entire dataset(80%): 0.15621892297850382 Dataset D1 R^2 on trained D1: 0.20816660043254143 .................................................. Pipeline #19: Score on D2: 0.1298447457696672 | D1-D2 diff: 1.9187740753697553 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=10, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20775845113212854 Holdout data R^2 trained on entire dataset(80%): 0.1559022755004824 Dataset D1 R^2 on trained D1: 0.20361896666246326 .................................................. Pipeline #20: Score on D2: 0.12920678313547074 | D1-D2 diff: 1.9531421401512696 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.1, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2048066531890862 Holdout data R^2 trained on entire dataset(80%): 0.15482538453294814 Dataset D1 R^2 on trained D1: 0.19792384766255366 .................................................. Pipeline #21: Score on D2: 0.12458186738566512 | D1-D2 diff: 2.0592186954900935 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18950254848272474 Holdout data R^2 trained on entire dataset(80%): 0.1536964121750335 Dataset D1 R^2 on trained D1: 0.18019663283758747 .................................................. Pipeline #22: Score on D2: 0.11656265419321932 | D1-D2 diff: 2.0765608680727246 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.1, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18375663628965166 Holdout data R^2 trained on entire dataset(80%): 0.145379885315149 Dataset D1 R^2 on trained D1: 0.17034272099992454 .................................................. Pipeline #23: Score on D2: 0.11612580458006849 | D1-D2 diff: 2.128042216715097 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17551497289743734 Holdout data R^2 trained on entire dataset(80%): 0.14244376844589623 Dataset D1 R^2 on trained D1: 0.16488752752126357 .................................................. Pipeline #24: Score on D2: 0.11474816475755845 | D1-D2 diff: 2.141435533282337 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17580886087554504 Holdout data R^2 trained on entire dataset(80%): 0.14359618908974237 Dataset D1 R^2 on trained D1: 0.16230139035679436 .................................................. Pipeline #25: Score on D2: 0.11092694835642869 | D1-D2 diff: 2.200643117055452 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16257412555187056 Holdout data R^2 trained on entire dataset(80%): 0.1320745901116701 Dataset D1 R^2 on trained D1: 0.1535654101378976 .................................................. Pipeline #26: Score on D2: 0.09250103783666885 | D1-D2 diff: 2.340606831143281 Pipeline steps: SelectPercentile(percentile=75), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=8, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11021381235392214 Holdout data R^2 trained on entire dataset(80%): 0.07831170737485782 Dataset D1 R^2 on trained D1: 0.12581959046369218 .................................................. Pipeline #27: Score on D2: 0.06227057742299569 | D1-D2 diff: 4.249168716500233 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(threshold=0.05), SelectPercentile(percentile=25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06392828404462847 Holdout data R^2 trained on entire dataset(80%): 0.036317638838094135 Dataset D1 R^2 on trained D1: 0.06533807408135484 .................................................. Pipeline #28: Score on D2: 0.061797054914313865 | D1-D2 diff: 4.313059898705593 Pipeline steps: VarianceThreshold(threshold=0.15), OverDominanceEncoder(), SelectPercentile(percentile=25), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06336796030069303 Holdout data R^2 trained on entire dataset(80%): 0.035837407215366435 Dataset D1 R^2 on trained D1: 0.0646867900718271 .................................................. Pipeline #29: Score on D2: 0.04276696508096345 | D1-D2 diff: 4.780533414880562 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=5), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.043774736078220466 Holdout data R^2 trained on entire dataset(80%): 0.018199105205561894 Dataset D1 R^2 on trained D1: 0.04468163787937918 .................................................. Pipeline #30: Score on D2: 0.03144508651338895 | D1-D2 diff: 5.059919073168936 Pipeline steps: RecessiveEncoder(), HeterosisEncoder(), SelectPercentile(percentile=15), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.6000000000000001, min_samples_leaf=4, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.031137357266389887 Holdout data R^2 trained on entire dataset(80%): 0.0062049425573226236 Dataset D1 R^2 on trained D1: 0.029919539077379764 .................................................. Pipeline #31: Score on D2: 0.02744978395077413 | D1-D2 diff: 5.837430543686509 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0304392513056454 Holdout data R^2 trained on entire dataset(80%): 0.011459662367232837 Dataset D1 R^2 on trained D1: 0.028311001944847702 .................................................. Pipeline #32: Score on D2: 0.025759847004414582 | D1-D2 diff: 12.625997144706586 Pipeline steps: SelectPercentile(percentile=70), RecessiveEncoder(), SelectPercentile(percentile=65), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03274866064948656 Holdout data R^2 trained on entire dataset(80%): 0.01333985602105403 Dataset D1 R^2 on trained D1: 0.025799196326337093 .................................................. Pipeline #33: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 31 - 8 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.17395 Best score on D1-D2 diff: 6.04150 Gen 2 - Best score on D2: 0.17395 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.17395 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.17540 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.17659 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.17659 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.17659 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.17659 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.18639 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.18639 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.010881912122125303 Entire dataset(80%) R^2 trained on data (80%): 0.0034648343508931756 Holdout R^2 (20%) trained on data (80%): -0.0003681810124283036 Dataset D1 R^2 on trained D1: 0.005404370035342332 Combined Dataset (100%) R^2 trained on combined data (100%): 0.00401649900756218 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.18639163862608932 | D1-D2 diff: 1.480814685215423 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.38580985493522046 Holdout data R^2 trained on entire dataset(80%): 0.22761206960617875 Dataset D1 R^2 on trained D1: 0.3943599422656674 .................................................. Pipeline #2: Score on D2: 0.18227908202008902 | D1-D2 diff: 1.533656533387803 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.35730336972181154 Holdout data R^2 trained on entire dataset(80%): 0.22472665589153962 Dataset D1 R^2 on trained D1: 0.3630329350916972 .................................................. Pipeline #3: Score on D2: 0.17658806494693757 | D1-D2 diff: 1.5998187994984872 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=14, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3205672908673103 Holdout data R^2 trained on entire dataset(80%): 0.21695651117515535 Dataset D1 R^2 on trained D1: 0.3292450976524999 .................................................. Pipeline #4: Score on D2: 0.17395088669022452 | D1-D2 diff: 1.6494047031713897 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30030227490951056 Holdout data R^2 trained on entire dataset(80%): 0.2099211137835849 Dataset D1 R^2 on trained D1: 0.3090620040657446 .................................................. Pipeline #5: Score on D2: 0.1720013940966041 | D1-D2 diff: 1.6615155597041975 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=8, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2925332032485358 Holdout data R^2 trained on entire dataset(80%): 0.20823367230033052 Dataset D1 R^2 on trained D1: 0.30321605100263693 .................................................. Pipeline #6: Score on D2: 0.17137394521019822 | D1-D2 diff: 1.6672929339015998 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.1), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29592867909619514 Holdout data R^2 trained on entire dataset(80%): 0.21533013148918445 Dataset D1 R^2 on trained D1: 0.300779333901815 .................................................. Pipeline #7: Score on D2: 0.17085191330022031 | D1-D2 diff: 1.711490171980463 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=13, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2824624406136391 Holdout data R^2 trained on entire dataset(80%): 0.20577172200076754 Dataset D1 R^2 on trained D1: 0.2873992521426709 .................................................. Pipeline #8: Score on D2: 0.16941464977270904 | D1-D2 diff: 1.8145542949707147 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2610430546230015 Holdout data R^2 trained on entire dataset(80%): 0.1990651333068939 Dataset D1 R^2 on trained D1: 0.2616548271335769 .................................................. Pipeline #9: Score on D2: 0.16093234729680994 | D1-D2 diff: 1.8308521387370265 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24717377637983673 Holdout data R^2 trained on entire dataset(80%): 0.1994937752155983 Dataset D1 R^2 on trained D1: 0.24993171335463749 .................................................. Pipeline #10: Score on D2: 0.14556761990302414 | D1-D2 diff: 1.9247903726926812 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), SelectPercentile(percentile=50), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20689522200051513 Holdout data R^2 trained on entire dataset(80%): 0.18229219439618227 Dataset D1 R^2 on trained D1: 0.21842377513649314 .................................................. Pipeline #11: Score on D2: 0.1305134266166087 | D1-D2 diff: 1.9720778786344932 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.05, min_samples_leaf=18, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2088340406326219 Holdout data R^2 trained on entire dataset(80%): 0.16575840447976675 Dataset D1 R^2 on trained D1: 0.19662899775602338 .................................................. Pipeline #12: Score on D2: 0.12155911352590054 | D1-D2 diff: 2.190625277846209 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.1), HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17582922712867288 Holdout data R^2 trained on entire dataset(80%): 0.14892167185738026 Dataset D1 R^2 on trained D1: 0.16498289306409608 .................................................. Pipeline #13: Score on D2: 0.10795799354228952 | D1-D2 diff: 2.38073901191136 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=7, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11694082388590488 Holdout data R^2 trained on entire dataset(80%): 0.09768198810136164 Dataset D1 R^2 on trained D1: 0.13908611022682982 .................................................. Pipeline #14: Score on D2: 0.06350546556460568 | D1-D2 diff: 6.041504852102362 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(threshold=0.25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06554664495483531 Holdout data R^2 trained on entire dataset(80%): 0.05319136023019566 Dataset D1 R^2 on trained D1: 0.06275484668166753 .................................................. Pipeline #15: Score on D2: 0.0628382362068044 | D1-D2 diff: 7.214415437384198 Pipeline steps: VarianceThreshold(threshold=0.3), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06454065096143058 Holdout data R^2 trained on entire dataset(80%): 0.05157445847022135 Dataset D1 R^2 on trained D1: 0.062469092549292005 .................................................. Pipeline #16: Score on D2: 0.050760506072863865 | D1-D2 diff: 7.235377011626254 Pipeline steps: SelectPercentile(percentile=40), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04509272762794103 Holdout data R^2 trained on entire dataset(80%): 0.029104251220828492 Dataset D1 R^2 on trained D1: 0.051125390507274826 .................................................. Pipeline #17: Score on D2: 0.0028375015596076025 | D1-D2 diff: 7.509874154250898 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0034583061816184646 Holdout data R^2 trained on entire dataset(80%): 0.002883248334442867 Dataset D1 R^2 on trained D1: 0.002523111097343933 .................................................. Pipeline #18: Score on D2: 0.0028375015596073805 | D1-D2 diff: 7.509874154251561 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0034583061816184646 Holdout data R^2 trained on entire dataset(80%): 0.002883248334442867 Dataset D1 R^2 on trained D1: 0.002523111097343822 .................................................. Pipeline #19: Score on D2: 0.0027541238753181485 | D1-D2 diff: 8.064630386685376 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), SelectPercentile(percentile=90), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.003445248612407026 Holdout data R^2 trained on entire dataset(80%): 0.0029800146173696307 Dataset D1 R^2 on trained D1: 0.0025177158972675695 .................................................. Pipeline #20: Score on D2: 0.0021861808588798937 | D1-D2 diff: 8.736253749731697 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), SelectPercentile(percentile=90), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0028133476296647864 Holdout data R^2 trained on entire dataset(80%): 0.001158105168927226 Dataset D1 R^2 on trained D1: 0.0020145090273943067 .................................................. Pipeline #21: Score on D2: 0.001977809134390607 | D1-D2 diff: 9.138866943984493 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002821971938397616 Holdout data R^2 trained on entire dataset(80%): 0.0010884776632912319 Dataset D1 R^2 on trained D1: 0.0021211699894748692 .................................................. Pipeline #22: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 31 - 9 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.17204 Best score on D1-D2 diff: 6.35370 Gen 2 - Best score on D2: 0.17309 Best score on D1-D2 diff: 9.13887 Gen 3 - Best score on D2: 0.17621 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.17621 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.17621 Best score on D1-D2 diff: 20.06045 Gen 6 - Best score on D2: 0.17621 Best score on D1-D2 diff: 20.06045 Gen 7 - Best score on D2: 0.17621 Best score on D1-D2 diff: 20.06045 Gen 8 - Best score on D2: 0.17621 Best score on D1-D2 diff: 20.06045 Gen 9 - Best score on D2: 0.17621 Best score on D1-D2 diff: 20.06045 Gen 10 - Best score on D2: 0.17621 Best score on D1-D2 diff: 20.06045 Gen 11 - Best score on D2: 0.17621 Best score on D1-D2 diff: 20.06045 Gen 12 - Best score on D2: 0.17826 Best score on D1-D2 diff: 20.06045 Gen 13 - Best score on D2: 0.17826 Best score on D1-D2 diff: 20.06045 Gen 14 - Best score on D2: 0.18069 Best score on D1-D2 diff: 20.06045 Gen 15 - Best score on D2: 0.18069 Best score on D1-D2 diff: 20.06045 Gen 16 - Best score on D2: 0.18083 Best score on D1-D2 diff: 20.06045 Gen 17 - Best score on D2: 0.18083 Best score on D1-D2 diff: 20.06045 Gen 18 - Best score on D2: 0.18083 Best score on D1-D2 diff: 20.06045 Gen 19 - Best score on D2: 0.18259 Best score on D1-D2 diff: 20.06045 Gen 20 - Best score on D2: 0.18259 Best score on D1-D2 diff: 20.06045 Gen 21 - Best score on D2: 0.18259 Best score on D1-D2 diff: 20.06045 Gen 22 - Best score on D2: 0.18259 Best score on D1-D2 diff: 20.06045 Gen 23 - Best score on D2: 0.18259 Best score on D1-D2 diff: 20.06045 Gen 24 - Best score on D2: 0.18259 Best score on D1-D2 diff: 20.06045 Gen 25 - Best score on D2: 0.18259 Best score on D1-D2 diff: 20.06045 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.008503172753477228 Entire dataset(80%) R^2 trained on data (80%): 0.004241391834748476 Holdout R^2 (20%) trained on data (80%): -0.002858968131327577 Dataset D1 R^2 on trained D1: 0.005725550136308266 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004309333642957491 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.18258749327951196 | D1-D2 diff: 1.5637961781813325 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34827689283235186 Holdout data R^2 trained on entire dataset(80%): 0.22893083228654743 Dataset D1 R^2 on trained D1: 0.3498041016176464 .................................................. Pipeline #2: Score on D2: 0.17730924400855363 | D1-D2 diff: 1.649489568785433 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=5, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3072352057227644 Holdout data R^2 trained on entire dataset(80%): 0.21959821387138145 Dataset D1 R^2 on trained D1: 0.3123925578664687 .................................................. Pipeline #3: Score on D2: 0.17442510630913588 | D1-D2 diff: 1.662353414237275 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2947551386125027 Holdout data R^2 trained on entire dataset(80%): 0.2158009085655237 Dataset D1 R^2 on trained D1: 0.30537542542627016 .................................................. Pipeline #4: Score on D2: 0.17209798819863065 | D1-D2 diff: 1.6901289990128565 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=20, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28705509377516814 Holdout data R^2 trained on entire dataset(80%): 0.21129579073149185 Dataset D1 R^2 on trained D1: 0.2946500399249783 .................................................. Pipeline #5: Score on D2: 0.1697949356505286 | D1-D2 diff: 1.7338427335868811 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2762796760191698 Holdout data R^2 trained on entire dataset(80%): 0.21093779451379802 Dataset D1 R^2 on trained D1: 0.28044742513357623 .................................................. Pipeline #6: Score on D2: 0.1694757495530138 | D1-D2 diff: 1.7348655607319425 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27677813893685155 Holdout data R^2 trained on entire dataset(80%): 0.2091003391803984 Dataset D1 R^2 on trained D1: 0.2798675195389245 .................................................. Pipeline #7: Score on D2: 0.1674399308667377 | D1-D2 diff: 1.7619091834702407 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2632538599659211 Holdout data R^2 trained on entire dataset(80%): 0.20553322103821525 Dataset D1 R^2 on trained D1: 0.2712085226702484 .................................................. Pipeline #8: Score on D2: 0.16707766080062603 | D1-D2 diff: 1.762013692177517 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=18, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26602336388952863 Holdout data R^2 trained on entire dataset(80%): 0.20297919927711772 Dataset D1 R^2 on trained D1: 0.2708216358678719 .................................................. Pipeline #9: Score on D2: 0.1667004701421676 | D1-D2 diff: 1.7802864985777178 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=12, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26051310897089996 Holdout data R^2 trained on entire dataset(80%): 0.2015527906366542 Dataset D1 R^2 on trained D1: 0.2662502736340825 .................................................. Pipeline #10: Score on D2: 0.16573096881250482 | D1-D2 diff: 1.7913750825433616 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=12, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25857315706105544 Holdout data R^2 trained on entire dataset(80%): 0.20378270457996073 Dataset D1 R^2 on trained D1: 0.2628387170984101 .................................................. Pipeline #11: Score on D2: 0.16556081882056017 | D1-D2 diff: 1.7914994304327114 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25746065423239084 Holdout data R^2 trained on entire dataset(80%): 0.20311181201754958 Dataset D1 R^2 on trained D1: 0.2626416089368133 .................................................. Pipeline #12: Score on D2: 0.1638646687590256 | D1-D2 diff: 1.900268870658431 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24066820770148256 Holdout data R^2 trained on entire dataset(80%): 0.1979856918827534 Dataset D1 R^2 on trained D1: 0.24055485351092099 .................................................. Pipeline #13: Score on D2: 0.15988120662152405 | D1-D2 diff: 1.9197999811486004 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23339464005227184 Holdout data R^2 trained on entire dataset(80%): 0.193526966863392 Dataset D1 R^2 on trained D1: 0.23349785952896862 .................................................. Pipeline #14: Score on D2: 0.1597145060992632 | D1-D2 diff: 1.935006711615971 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22889526852291597 Holdout data R^2 trained on entire dataset(80%): 0.19454163955824333 Dataset D1 R^2 on trained D1: 0.2310441567719811 .................................................. Pipeline #15: Score on D2: 0.15814069262526154 | D1-D2 diff: 1.9589045728550705 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22920325301242528 Holdout data R^2 trained on entire dataset(80%): 0.19137435165614125 Dataset D1 R^2 on trained D1: 0.22605274879738646 .................................................. Pipeline #16: Score on D2: 0.15086831433464243 | D1-D2 diff: 1.9651360527376625 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22119879339742443 Holdout data R^2 trained on entire dataset(80%): 0.18492822816532672 Dataset D1 R^2 on trained D1: 0.21792305798279243 .................................................. Pipeline #17: Score on D2: 0.14823709547447317 | D1-D2 diff: 2.010531387086995 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20953288425455252 Holdout data R^2 trained on entire dataset(80%): 0.17921461321587073 Dataset D1 R^2 on trained D1: 0.20943782095579933 .................................................. Pipeline #18: Score on D2: 0.14730724639948645 | D1-D2 diff: 2.0217223539096616 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=14, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21602596577524835 Holdout data R^2 trained on entire dataset(80%): 0.17826056942562996 Dataset D1 R^2 on trained D1: 0.2071641087584034 .................................................. Pipeline #19: Score on D2: 0.14531868505047962 | D1-D2 diff: 2.0431638972409223 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20896562534115615 Holdout data R^2 trained on entire dataset(80%): 0.17792842776925866 Dataset D1 R^2 on trained D1: 0.20270220385625592 .................................................. Pipeline #20: Score on D2: 0.1449204093045955 | D1-D2 diff: 2.0504470229919227 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20841715035672448 Holdout data R^2 trained on entire dataset(80%): 0.17704086833402943 Dataset D1 R^2 on trained D1: 0.20149296361814617 .................................................. Pipeline #21: Score on D2: 0.14467779521637736 | D1-D2 diff: 2.0655375018561983 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20455256850908454 Holdout data R^2 trained on entire dataset(80%): 0.17819908810026952 Dataset D1 R^2 on trained D1: 0.19961513960840271 .................................................. Pipeline #22: Score on D2: 0.14003964587224682 | D1-D2 diff: 2.0837322706761587 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.2, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20029297299435278 Holdout data R^2 trained on entire dataset(80%): 0.17160060431943247 Dataset D1 R^2 on trained D1: 0.19308316500272793 .................................................. Pipeline #23: Score on D2: 0.13817278894021168 | D1-D2 diff: 2.146436159186668 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.2, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.194204089147479 Holdout data R^2 trained on entire dataset(80%): 0.16829923289307713 Dataset D1 R^2 on trained D1: 0.1852844152329618 .................................................. Pipeline #24: Score on D2: 0.13502596052768046 | D1-D2 diff: 2.1782205749978436 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=18, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18656629864475893 Holdout data R^2 trained on entire dataset(80%): 0.1665861171662475 Dataset D1 R^2 on trained D1: 0.17944739472025062 .................................................. Pipeline #25: Score on D2: 0.13236511028383136 | D1-D2 diff: 2.2139225262136857 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18641250260682973 Holdout data R^2 trained on entire dataset(80%): 0.16147422460240834 Dataset D1 R^2 on trained D1: 0.17398973427407427 .................................................. Pipeline #26: Score on D2: 0.13134902184343167 | D1-D2 diff: 2.224280054695632 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.1, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1868883337388304 Holdout data R^2 trained on entire dataset(80%): 0.16553694726382706 Dataset D1 R^2 on trained D1: 0.1722037316435272 .................................................. Pipeline #27: Score on D2: 0.12802157924112156 | D1-D2 diff: 2.23819759353562 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18196754861357078 Holdout data R^2 trained on entire dataset(80%): 0.16407956136506585 Dataset D1 R^2 on trained D1: 0.1678695584532247 .................................................. Pipeline #28: Score on D2: 0.12472192951310968 | D1-D2 diff: 2.254317346555365 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1753365311421996 Holdout data R^2 trained on entire dataset(80%): 0.15625204662658498 Dataset D1 R^2 on trained D1: 0.16344232536064207 .................................................. Pipeline #29: Score on D2: 0.12263517220259579 | D1-D2 diff: 2.2751254675400476 Pipeline steps: SelectPercentile(percentile=65), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=20, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17038236142584384 Holdout data R^2 trained on entire dataset(80%): 0.13929657236652815 Dataset D1 R^2 on trained D1: 0.15995834816545718 .................................................. Pipeline #30: Score on D2: 0.11183777119564942 | D1-D2 diff: 2.3005608465573566 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=20, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15936592133721994 Holdout data R^2 trained on entire dataset(80%): 0.1397665633616615 Dataset D1 R^2 on trained D1: 0.147537515300456 .................................................. Pipeline #31: Score on D2: 0.09089353375325648 | D1-D2 diff: 2.3106711400433997 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=20, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10993421594055475 Holdout data R^2 trained on entire dataset(80%): 0.07864010379559694 Dataset D1 R^2 on trained D1: 0.1259725527503056 .................................................. Pipeline #32: Score on D2: 0.07827545010941994 | D1-D2 diff: 2.6067014162308286 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RecessiveEncoder(), HeterosisEncoder(), RecessiveEncoder(), SelectPercentile(percentile=30), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09030649966711624 Holdout data R^2 trained on entire dataset(80%): 0.09044183783442106 Dataset D1 R^2 on trained D1: 0.09993427292190271 .................................................. Pipeline #33: Score on D2: 0.07726010141855977 | D1-D2 diff: 2.680794014295232 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), HeterosisEncoder(), SelectPercentile(percentile=25), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0879368106548638 Holdout data R^2 trained on entire dataset(80%): 0.09172511609783929 Dataset D1 R^2 on trained D1: 0.09662192325319952 .................................................. Pipeline #34: Score on D2: 0.0746788676126513 | D1-D2 diff: 3.3204798095126873 Pipeline steps: VarianceThreshold(), UnderDominanceEncoder(), SelectPercentile(percentile=35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0795831119631204 Holdout data R^2 trained on entire dataset(80%): 0.08013232247292079 Dataset D1 R^2 on trained D1: 0.08290501761906655 .................................................. Pipeline #35: Score on D2: 0.07269970200451181 | D1-D2 diff: 3.4119295819705777 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0772098833294651 Holdout data R^2 trained on entire dataset(80%): 0.08064041649273146 Dataset D1 R^2 on trained D1: 0.08007874027664841 .................................................. Pipeline #36: Score on D2: 0.071646813263056 | D1-D2 diff: 4.543157028535858 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0737106024210028 Holdout data R^2 trained on entire dataset(80%): 0.07751802516000061 Dataset D1 R^2 on trained D1: 0.07399411547665002 .................................................. Pipeline #37: Score on D2: 0.07134123647320656 | D1-D2 diff: 4.648354795816916 Pipeline steps: VarianceThreshold(threshold=0.2), UnderDominanceEncoder(), SelectPercentile(percentile=25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07292723239525589 Holdout data R^2 trained on entire dataset(80%): 0.07828031024534943 Dataset D1 R^2 on trained D1: 0.07348315488251622 .................................................. Pipeline #38: Score on D2: 0.06810625634838907 | D1-D2 diff: 6.353696994821701 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06889325074302621 Holdout data R^2 trained on entire dataset(80%): 0.06799463207853795 Dataset D1 R^2 on trained D1: 0.06871986857039003 .................................................. Pipeline #39: Score on D2: 0.005755505241651915 | D1-D2 diff: 7.8409028088274555 Pipeline steps: VarianceThreshold(threshold=0.1), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), SelectPercentile(percentile=80), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.025303868348470093 Holdout data R^2 trained on entire dataset(80%): 0.023690100896614608 Dataset D1 R^2 on trained D1: 0.005490938214418595 .................................................. Pipeline #40: Score on D2: 0.005639141559724781 | D1-D2 diff: 20.060452442882443 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), FeatureEncodingFrequencySelector(threshold=0.1), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.025318292634195605 Holdout data R^2 trained on entire dataset(80%): 0.02390435251409806 Dataset D1 R^2 on trained D1: 0.00563296655769685 .................................................. ************************************************************************************** Random Seed 32 - 0 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00000 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.00029 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00058 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.005874311881761685 Entire dataset(80%) R^2 trained on data (80%): 0.0039082518204752725 Holdout R^2 (20%) trained on data (80%): -0.001165662515316912 Dataset D1 R^2 on trained D1: 0.006828780271858292 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0034897346911435534 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.003224355638801968 | D1-D2 diff: 2.0840488253768954 Pipeline steps: VarianceThreshold(threshold=0.2), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=18, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06065062785272424 Holdout data R^2 trained on entire dataset(80%): -0.005601272566658766 Dataset D1 R^2 on trained D1: 0.05623565412303544 .................................................. Pipeline #2: Score on D2: 0.0005947533841992314 | D1-D2 diff: 4.4248570118992285 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=25), DecisionTreeRegressor(max_depth=2, min_samples_leaf=20, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0025360855898788337 Holdout data R^2 trained on entire dataset(80%): -0.0008173957317167968 Dataset D1 R^2 on trained D1: 0.0032033266280756534 .................................................. Pipeline #3: Score on D2: 0.0005803695349020366 | D1-D2 diff: 6.664227131597254 Pipeline steps: VarianceThreshold(threshold=0.3), SelectPercentile(percentile=50), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0012685735942515874 Holdout data R^2 trained on entire dataset(80%): -0.004112975494448001 Dataset D1 R^2 on trained D1: 0.0010873612220696494 .................................................. Pipeline #4: Score on D2: 0.0003651372031471256 | D1-D2 diff: 8.113629181900814 Pipeline steps: VarianceThreshold(threshold=0.3), SelectPercentile(percentile=25), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.000542925019066387 Holdout data R^2 trained on entire dataset(80%): -0.002873875653820912 Dataset D1 R^2 on trained D1: 0.0005958859648796944 .................................................. Pipeline #5: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: RecessiveEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.5, min_samples_leaf=9, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 32 - 1 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00000 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.00029 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00058 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.08730 Gen 2 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.89837 Gen 3 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 14 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 15 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 16 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 17 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 18 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 19 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 20 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 21 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.005993063067430304 Entire dataset(80%) R^2 trained on data (80%): 0.0039260609415343595 Holdout R^2 (20%) trained on data (80%): -8.631017338278646e-05 Dataset D1 R^2 on trained D1: 0.006836204647836519 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003773889910293926 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.09117193462839579 | D1-D2 diff: 3.0642050339264455 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), DecisionTreeRegressor(max_depth=2, min_samples_leaf=19, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0028929709279834626 Holdout data R^2 trained on entire dataset(80%): -0.005590332834574552 Dataset D1 R^2 on trained D1: 0.10251495502920749 .................................................. Pipeline #2: Score on D2: 0.06059405449330091 | D1-D2 diff: 3.083026855162319 Pipeline steps: VarianceThreshold(threshold=0.15), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.007808920059715563 Holdout data R^2 trained on entire dataset(80%): -0.010587728574183108 Dataset D1 R^2 on trained D1: 0.07166260545763081 .................................................. Pipeline #3: Score on D2: 0.007647135353972101 | D1-D2 diff: 3.5063021559144603 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=19, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0029933391496180795 Holdout data R^2 trained on entire dataset(80%): -0.004585310602283954 Dataset D1 R^2 on trained D1: 0.01426324425448855 .................................................. Pipeline #4: Score on D2: 0.0030444071021007435 | D1-D2 diff: 3.9922986439175494 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=17, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0011661240426398267 Holdout data R^2 trained on entire dataset(80%): -0.003986781013539442 Dataset D1 R^2 on trained D1: 0.0069808858857346445 .................................................. Pipeline #5: Score on D2: 0.0001970625527891734 | D1-D2 diff: 6.654971534637721 Pipeline steps: SelectPercentile(percentile=20), FeatureEncodingFrequencySelector(threshold=0.15), VarianceThreshold(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0004985749418819907 Holdout data R^2 trained on entire dataset(80%): -0.00028803571187441257 Dataset D1 R^2 on trained D1: 0.000706880583627223 .................................................. Pipeline #6: Score on D2: 8.532920604542582e-07 | D1-D2 diff: 16.976242685478304 Pipeline steps: SelectPercentile(percentile=20), SelectPercentile(percentile=15), DominantEncoder(), VarianceThreshold(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 3.4888357375417733e-05 Holdout data R^2 trained on entire dataset(80%): -0.0008494432352597059 Dataset D1 R^2 on trained D1: 1.2893492020160657e-05 .................................................. ************************************************************************************** Random Seed 32 - 2 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00000 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.00029 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00058 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.08730 Gen 2 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.89837 Gen 3 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 14 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 15 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 16 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 17 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 18 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 19 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 20 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 21 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 2 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 3 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.10063 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.10239 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 12 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.006840078833329821 Entire dataset(80%) R^2 trained on data (80%): 0.003102336492762703 Holdout R^2 (20%) trained on data (80%): -0.002446827091418191 Dataset D1 R^2 on trained D1: 0.0057273625441178755 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0027875084335661215 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.10513737230661868 | D1-D2 diff: 1.7303282984901622 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(threshold=0.25), RecessiveEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21284624429071075 Holdout data R^2 trained on entire dataset(80%): 0.13678141188252224 Dataset D1 R^2 on trained D1: 0.21669158054570425 .................................................. Pipeline #2: Score on D2: 0.10286025414411548 | D1-D2 diff: 1.7373390543015395 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21439754234182073 Holdout data R^2 trained on entire dataset(80%): 0.1331204675126656 Dataset D1 R^2 on trained D1: 0.21262469521252214 .................................................. Pipeline #3: Score on D2: 0.1013397454638092 | D1-D2 diff: 1.7918991552889014 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=19, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19816198340511126 Holdout data R^2 trained on entire dataset(80%): 0.13132473619894514 Dataset D1 R^2 on trained D1: 0.1983339400326667 .................................................. Pipeline #4: Score on D2: 0.09173058319925509 | D1-D2 diff: 3.059640755508997 Pipeline steps: VarianceThreshold(threshold=0.1), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09183813600152202 Holdout data R^2 trained on entire dataset(80%): 0.14781162017060057 Dataset D1 R^2 on trained D1: 0.1031414398880598 .................................................. Pipeline #5: Score on D2: 0.06061125558446934 | D1-D2 diff: 3.0654347326909064 Pipeline steps: VarianceThreshold(threshold=0.3), DecisionTreeRegressor(max_depth=3, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.008973468126477702 Holdout data R^2 trained on entire dataset(80%): 0.0009922387693424017 Dataset D1 R^2 on trained D1: 0.0719360859294167 .................................................. Pipeline #6: Score on D2: 0.008629233864579589 | D1-D2 diff: 3.573032863791457 Pipeline steps: DominantEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00241399160480682 Holdout data R^2 trained on entire dataset(80%): -0.002582971258380473 Dataset D1 R^2 on trained D1: 0.0147647620713387 .................................................. Pipeline #7: Score on D2: 7.97600625898065e-05 | D1-D2 diff: 5.118863770529705 Pipeline steps: SelectPercentile(percentile=50), SelectPercentile(percentile=25), SelectPercentile(percentile=50), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.000928120411878619 Holdout data R^2 trained on entire dataset(80%): 0.0003713208302889015 Dataset D1 R^2 on trained D1: 0.0015362440457573623 .................................................. Pipeline #8: Score on D2: 8.532920604542582e-07 | D1-D2 diff: 16.976242685478304 Pipeline steps: SelectPercentile(percentile=50), SelectPercentile(percentile=25), SelectPercentile(percentile=30), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 3.4888357375417733e-05 Holdout data R^2 trained on entire dataset(80%): -0.0008494432352597059 Dataset D1 R^2 on trained D1: 1.2893492020160657e-05 .................................................. ************************************************************************************** Random Seed 32 - 3 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00000 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.00029 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00058 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.08730 Gen 2 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.89837 Gen 3 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 14 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 15 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 16 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 17 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 18 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 19 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 20 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 21 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 2 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 3 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.10063 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.10239 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 12 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.11435 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.12839 Best score on D1-D2 diff: 5.11886 Gen 3 - Best score on D2: 0.12839 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.12840 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.13233 Best score on D1-D2 diff: 16.97624 Gen 6 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 7 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 8 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 9 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 12 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 13 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.00442669005502494 Entire dataset(80%) R^2 trained on data (80%): 0.004598244592505796 Holdout R^2 (20%) trained on data (80%): -0.001610209792513917 Dataset D1 R^2 on trained D1: 0.006265775086535363 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004370638206811606 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.13821465690488088 | D1-D2 diff: 1.855536027851219 Pipeline steps: SelectPercentile(percentile=65), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=11, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19494691192795233 Holdout data R^2 trained on entire dataset(80%): 0.06812170569354337 Dataset D1 R^2 on trained D1: 0.2225719111072536 .................................................. Pipeline #2: Score on D2: 0.13233129793634923 | D1-D2 diff: 2.078624830247012 Pipeline steps: SelectPercentile(percentile=65), HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16236187498300858 Holdout data R^2 trained on entire dataset(80%): 0.07103389310603958 Dataset D1 R^2 on trained D1: 0.18589807987148355 .................................................. Pipeline #3: Score on D2: 0.11084143412602698 | D1-D2 diff: 2.2035298328141257 Pipeline steps: SelectPercentile(percentile=65), HeterosisEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=11, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1248906849485264 Holdout data R^2 trained on entire dataset(80%): 0.048422606282932046 Dataset D1 R^2 on trained D1: 0.15325690194762276 .................................................. Pipeline #4: Score on D2: 0.09022916964824246 | D1-D2 diff: 2.740835047742412 Pipeline steps: VarianceThreshold(threshold=0.35), FeatureEncodingFrequencySelector(threshold=0.1), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=11, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.006768249767555412 Holdout data R^2 trained on entire dataset(80%): 0.004632643559434713 Dataset D1 R^2 on trained D1: 0.10794936085034401 .................................................. Pipeline #5: Score on D2: 0.01196115263064168 | D1-D2 diff: 3.4065463866306036 Pipeline steps: VarianceThreshold(threshold=0.35), DominantEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=11, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04795387648700011 Holdout data R^2 trained on entire dataset(80%): 0.03559024954851897 Dataset D1 R^2 on trained D1: 0.01938694448497602 .................................................. Pipeline #6: Score on D2: 0.003537851128072589 | D1-D2 diff: 5.192262147867333 Pipeline steps: SelectPercentile(percentile=60), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.1), VarianceThreshold(threshold=0.25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00409802668903525 Holdout data R^2 trained on entire dataset(80%): 0.003597749479232526 Dataset D1 R^2 on trained D1: 0.002161993311000976 .................................................. Pipeline #7: Score on D2: 0.003307330932609487 | D1-D2 diff: 6.702809726918833 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0034713140940212384 Holdout data R^2 trained on entire dataset(80%): 0.0051007421877363734 Dataset D1 R^2 on trained D1: 0.0028119121872870467 .................................................. Pipeline #8: Score on D2: 3.3338264600057954e-05 | D1-D2 diff: 13.402122998261358 Pipeline steps: SelectPercentile(percentile=5), DominantEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 3.270712007419352e-05 Holdout data R^2 trained on entire dataset(80%): -0.0008506306079107961 Dataset D1 R^2 on trained D1: 2.3422681457718753e-06 .................................................. Pipeline #9: Score on D2: 2.277255174321091e-05 | D1-D2 diff: 16.979620735520363 Pipeline steps: SelectPercentile(percentile=5), DominantEncoder(), RandomForestRegressor(max_features=0.8, min_samples_leaf=20, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 3.270712007419352e-05 Holdout data R^2 trained on entire dataset(80%): -0.0008506306079107961 Dataset D1 R^2 on trained D1: 1.07419303866374e-05 .................................................. ************************************************************************************** Random Seed 32 - 4 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00000 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.00029 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00058 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.08730 Gen 2 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.89837 Gen 3 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 14 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 15 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 16 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 17 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 18 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 19 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 20 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 21 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 2 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 3 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.10063 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.10239 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 12 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.11435 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.12839 Best score on D1-D2 diff: 5.11886 Gen 3 - Best score on D2: 0.12839 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.12840 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.13233 Best score on D1-D2 diff: 16.97624 Gen 6 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 7 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 8 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 9 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 12 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 13 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 1 - Best score on D2: 0.14656 Best score on D1-D2 diff: 7.98775 Gen 2 - Best score on D2: 0.14969 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14969 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0049195834591675425 Entire dataset(80%) R^2 trained on data (80%): 0.003839654384253266 Holdout R^2 (20%) trained on data (80%): -0.0005033465393102787 Dataset D1 R^2 on trained D1: 0.005781673078745886 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0037191940225469455 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.152407565039685 | D1-D2 diff: 1.9535136026898134 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=85), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23874219907207628 Holdout data R^2 trained on entire dataset(80%): 0.19049445904101614 Dataset D1 R^2 on trained D1: 0.22107237800224366 .................................................. Pipeline #2: Score on D2: 0.1414728399902132 | D1-D2 diff: 1.9596349138299092 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=20, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2357890254537247 Holdout data R^2 trained on entire dataset(80%): 0.18252726530001728 Dataset D1 R^2 on trained D1: 0.20928371152612324 .................................................. Pipeline #3: Score on D2: 0.12040990634078375 | D1-D2 diff: 2.1809965634774264 Pipeline steps: SelectPercentile(percentile=50), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19804652950955448 Holdout data R^2 trained on entire dataset(80%): 0.13104656049501562 Dataset D1 R^2 on trained D1: 0.16460561224005865 .................................................. Pipeline #4: Score on D2: 0.11976262196748277 | D1-D2 diff: 3.033022502842366 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=17, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1394068614570192 Holdout data R^2 trained on entire dataset(80%): 0.15106289684280194 Dataset D1 R^2 on trained D1: 0.13157935625660444 .................................................. Pipeline #5: Score on D2: 0.09598092202071629 | D1-D2 diff: 3.502230765733847 Pipeline steps: VarianceThreshold(threshold=0.35), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10485482225201337 Holdout data R^2 trained on entire dataset(80%): 0.13563317610154568 Dataset D1 R^2 on trained D1: 0.10262784987126572 .................................................. Pipeline #6: Score on D2: 0.06863785617951579 | D1-D2 diff: 3.5367287959478477 Pipeline steps: SelectPercentile(percentile=80), SelectPercentile(percentile=50), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.107146185858639 Holdout data R^2 trained on entire dataset(80%): 0.1136583795854289 Dataset D1 R^2 on trained D1: 0.0750292115568354 .................................................. Pipeline #7: Score on D2: 0.008961514707127916 | D1-D2 diff: 5.011385174106339 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.012733117223311163 Holdout data R^2 trained on entire dataset(80%): 0.0006718421567475374 Dataset D1 R^2 on trained D1: 0.010547024266175598 .................................................. Pipeline #8: Score on D2: 0.008438057978760738 | D1-D2 diff: 6.119359131169377 Pipeline steps: VarianceThreshold(threshold=0.3), HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011707456171234143 Holdout data R^2 trained on entire dataset(80%): 0.0004266386978513115 Dataset D1 R^2 on trained D1: 0.00915120033764727 .................................................. Pipeline #9: Score on D2: 0.00744876541830386 | D1-D2 diff: 7.987749194060049 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(threshold=0.25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.010056977191961969 Holdout data R^2 trained on entire dataset(80%): 0.0019820725442886156 Dataset D1 R^2 on trained D1: 0.007694407245778301 .................................................. Pipeline #10: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: DominantEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 32 - 5 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00000 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.00029 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00058 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.08730 Gen 2 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.89837 Gen 3 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 14 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 15 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 16 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 17 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 18 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 19 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 20 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 21 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 2 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 3 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.10063 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.10239 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 12 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.11435 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.12839 Best score on D1-D2 diff: 5.11886 Gen 3 - Best score on D2: 0.12839 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.12840 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.13233 Best score on D1-D2 diff: 16.97624 Gen 6 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 7 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 8 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 9 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 12 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 13 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 1 - Best score on D2: 0.14656 Best score on D1-D2 diff: 7.98775 Gen 2 - Best score on D2: 0.14969 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14969 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.16390 Best score on D1-D2 diff: 3.72276 Gen 2 - Best score on D2: 0.16463 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.16463 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.16463 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.16463 Best score on D1-D2 diff: 31.24909 Gen 6 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 7 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 8 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 9 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 10 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 11 - Best score on D2: 0.16823 Best score on D1-D2 diff: 31.24909 Gen 12 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 13 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 14 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.011514539648215916 Entire dataset(80%) R^2 trained on data (80%): 0.0040766797968903035 Holdout R^2 (20%) trained on data (80%): -0.0021421727831527626 Dataset D1 R^2 on trained D1: 0.009012778723174186 Combined Dataset (100%) R^2 trained on combined data (100%): 0.003590247307309613 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.16943957809532162 | D1-D2 diff: 1.40748681353007 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), FeatureEncodingFrequencySelector(threshold=0.0), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=4, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4263519634582362 Holdout data R^2 trained on entire dataset(80%): 0.21099880993032383 Dataset D1 R^2 on trained D1: 0.42425321184058284 .................................................. Pipeline #2: Score on D2: 0.16768020070286838 | D1-D2 diff: 1.4185365572330788 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=3, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.42361379366659035 Holdout data R^2 trained on entire dataset(80%): 0.21048199085689823 Dataset D1 R^2 on trained D1: 0.4146466000215504 .................................................. Pipeline #3: Score on D2: 0.16650539571546352 | D1-D2 diff: 1.5513258590799859 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), HeterosisEncoder(), VarianceThreshold(threshold=0.15), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.45, min_samples_leaf=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3457360956564013 Holdout data R^2 trained on entire dataset(80%): 0.20109322682883457 Dataset D1 R^2 on trained D1: 0.3391638595992639 .................................................. Pipeline #4: Score on D2: 0.1649646514397981 | D1-D2 diff: 1.551330331517671 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), VarianceThreshold(threshold=0.15), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.45, min_samples_leaf=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34582339504184567 Holdout data R^2 trained on entire dataset(80%): 0.2079449331262041 Dataset D1 R^2 on trained D1: 0.3376211242560212 .................................................. Pipeline #5: Score on D2: 0.16401308361413514 | D1-D2 diff: 1.5815414103416963 Pipeline steps: OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.5, min_samples_leaf=17, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.333779180963622 Holdout data R^2 trained on entire dataset(80%): 0.188717119646586 Dataset D1 R^2 on trained D1: 0.32385023425568515 .................................................. Pipeline #6: Score on D2: 0.1637220526772143 | D1-D2 diff: 1.5918978315776913 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=9, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33157794696525744 Holdout data R^2 trained on entire dataset(80%): 0.19942382421672078 Dataset D1 R^2 on trained D1: 0.31944020263607853 .................................................. Pipeline #7: Score on D2: 0.16344953158427733 | D1-D2 diff: 1.7031611398168487 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), VarianceThreshold(threshold=0.15), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.4, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2949406718095908 Holdout data R^2 trained on entire dataset(80%): 0.20431956386974504 Dataset D1 R^2 on trained D1: 0.2822934717867629 .................................................. Pipeline #8: Score on D2: 0.16278821612155647 | D1-D2 diff: 1.7099698588359826 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), VarianceThreshold(threshold=0.15), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.4, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2935526195185938 Holdout data R^2 trained on entire dataset(80%): 0.20390049605700844 Dataset D1 R^2 on trained D1: 0.27975059127468616 .................................................. Pipeline #9: Score on D2: 0.1606546228855067 | D1-D2 diff: 1.741331939077165 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28110135627137467 Holdout data R^2 trained on entire dataset(80%): 0.2033756002224807 Dataset D1 R^2 on trained D1: 0.26941575929504236 .................................................. Pipeline #10: Score on D2: 0.16011116009177617 | D1-D2 diff: 2.0328133204729517 Pipeline steps: SelectPercentile(percentile=65), HeterosisEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21858963424100974 Holdout data R^2 trained on entire dataset(80%): 0.16437117777846355 Dataset D1 R^2 on trained D1: 0.2186723656341396 .................................................. Pipeline #11: Score on D2: 0.14016873622102177 | D1-D2 diff: 3.0606129407283733 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), DecisionTreeRegressor(max_depth=4, min_samples_leaf=3, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1355649645433764 Holdout data R^2 trained on entire dataset(80%): 0.1434401093368225 Dataset D1 R^2 on trained D1: 0.15156510145744306 .................................................. Pipeline #12: Score on D2: 0.10346570471762317 | D1-D2 diff: 3.7227571540345425 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=80), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=3, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10689850640325038 Holdout data R^2 trained on entire dataset(80%): 0.1145764153484109 Dataset D1 R^2 on trained D1: 0.1086721483691303 .................................................. Pipeline #13: Score on D2: 0.03252880836013505 | D1-D2 diff: 6.284298641740724 Pipeline steps: DominantEncoder(), VarianceThreshold(threshold=0.1), HeterosisEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=6, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03300866401244773 Holdout data R^2 trained on entire dataset(80%): 0.027183358858805406 Dataset D1 R^2 on trained D1: 0.03316997768815855 .................................................. Pipeline #14: Score on D2: 0.01196051503069362 | D1-D2 diff: 7.155035948421792 Pipeline steps: DominantEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=9, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.012177163125235313 Holdout data R^2 trained on entire dataset(80%): 0.022271867387798294 Dataset D1 R^2 on trained D1: 0.012342066139210184 .................................................. Pipeline #15: Score on D2: 0.000222312371553679 | D1-D2 diff: 8.471543113565131 Pipeline steps: SelectPercentile(percentile=10), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0010222958334888954 Holdout data R^2 trained on entire dataset(80%): -0.0004123862719656035 Dataset D1 R^2 on trained D1: 0.00041646796128680386 .................................................. Pipeline #16: Score on D2: 8.877873152068894e-06 | D1-D2 diff: 18.1708337567683 Pipeline steps: SelectPercentile(percentile=10), SelectPercentile(percentile=10), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0009889737701765933 Holdout data R^2 trained on entire dataset(80%): -0.00023416907214302007 Dataset D1 R^2 on trained D1: 1.805064474924567e-05 .................................................. Pipeline #17: Score on D2: -3.7614835004529112e-06 | D1-D2 diff: 31.2490886587804 Pipeline steps: SelectPercentile(percentile=65), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), RandomForestRegressor(max_features=0.55, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0001641528022418015 Holdout data R^2 trained on entire dataset(80%): -0.0007130745719328146 Dataset D1 R^2 on trained D1: -4.810181827519244e-06 .................................................. ************************************************************************************** Random Seed 32 - 6 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00000 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.00029 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00058 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.08730 Gen 2 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.89837 Gen 3 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 14 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 15 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 16 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 17 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 18 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 19 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 20 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 21 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 2 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 3 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.10063 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.10239 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 12 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.11435 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.12839 Best score on D1-D2 diff: 5.11886 Gen 3 - Best score on D2: 0.12839 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.12840 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.13233 Best score on D1-D2 diff: 16.97624 Gen 6 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 7 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 8 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 9 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 12 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 13 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 1 - Best score on D2: 0.14656 Best score on D1-D2 diff: 7.98775 Gen 2 - Best score on D2: 0.14969 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14969 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.16390 Best score on D1-D2 diff: 3.72276 Gen 2 - Best score on D2: 0.16463 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.16463 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.16463 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.16463 Best score on D1-D2 diff: 31.24909 Gen 6 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 7 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 8 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 9 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 10 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 11 - Best score on D2: 0.16823 Best score on D1-D2 diff: 31.24909 Gen 12 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 13 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 14 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 1 - Best score on D2: 0.16861 Best score on D1-D2 diff: 3.72276 Gen 2 - Best score on D2: 0.17263 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.17422 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.17422 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.17422 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.17422 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.17422 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.17422 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.012807590683646009 Entire dataset(80%) R^2 trained on data (80%): 0.0029568007077094283 Holdout R^2 (20%) trained on data (80%): -0.0038163417488112916 Dataset D1 R^2 on trained D1: 0.008013801741674986 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0023605058234209553 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.17422268204903646 | D1-D2 diff: 1.583718392905952 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=11, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33690766733014343 Holdout data R^2 trained on entire dataset(80%): 0.21778028326440546 Dataset D1 R^2 on trained D1: 0.3331827932281739 .................................................. Pipeline #2: Score on D2: 0.17291872220309068 | D1-D2 diff: 1.5839851065283057 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=11, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33478806112804094 Holdout data R^2 trained on entire dataset(80%): 0.21475557226873543 Dataset D1 R^2 on trained D1: 0.3317717967192335 .................................................. Pipeline #3: Score on D2: 0.1717377324978493 | D1-D2 diff: 1.6701694501668798 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=11, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3115910160260317 Holdout data R^2 trained on entire dataset(80%): 0.2135390277969743 Dataset D1 R^2 on trained D1: 0.3002539271986189 .................................................. Pipeline #4: Score on D2: 0.17005868893817622 | D1-D2 diff: 1.7499132100556731 Pipeline steps: VarianceThreshold(threshold=0.15), VarianceThreshold(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2879414394506695 Holdout data R^2 trained on entire dataset(80%): 0.213315440329382 Dataset D1 R^2 on trained D1: 0.2767020836958479 .................................................. Pipeline #5: Score on D2: 0.1686050214752236 | D1-D2 diff: 1.7599056427820525 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2852017575771554 Holdout data R^2 trained on entire dataset(80%): 0.21246150847155088 Dataset D1 R^2 on trained D1: 0.272846956597939 .................................................. Pipeline #6: Score on D2: 0.16719688369119634 | D1-D2 diff: 1.8012022338479978 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=19, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27476605740095217 Holdout data R^2 trained on entire dataset(80%): 0.2083841961389853 Dataset D1 R^2 on trained D1: 0.2622026778042713 .................................................. Pipeline #7: Score on D2: 0.16475354994466163 | D1-D2 diff: 1.8608188338356872 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=19, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2603494136984189 Holdout data R^2 trained on entire dataset(80%): 0.21176205596517395 Dataset D1 R^2 on trained D1: 0.2481569254806797 .................................................. Pipeline #8: Score on D2: 0.16059189095669701 | D1-D2 diff: 1.8644348611921073 Pipeline steps: SelectPercentile(percentile=85), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=11, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25031663767811796 Holdout data R^2 trained on entire dataset(80%): 0.18424744575230367 Dataset D1 R^2 on trained D1: 0.24335011092105896 .................................................. Pipeline #9: Score on D2: 0.16039344373364517 | D1-D2 diff: 1.8864369273166353 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=85), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25549072977299436 Holdout data R^2 trained on entire dataset(80%): 0.21416101929945897 Dataset D1 R^2 on trained D1: 0.23935775342656807 .................................................. Pipeline #10: Score on D2: 0.15989493513120745 | D1-D2 diff: 1.9536861452055443 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24204385076539925 Holdout data R^2 trained on entire dataset(80%): 0.20208355930838673 Dataset D1 R^2 on trained D1: 0.2285354943922464 .................................................. Pipeline #11: Score on D2: 0.15301362882851255 | D1-D2 diff: 2.009157038421473 Pipeline steps: SelectPercentile(percentile=65), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21807607039859012 Holdout data R^2 trained on entire dataset(80%): 0.17960256561304844 Dataset D1 R^2 on trained D1: 0.21438198178051615 .................................................. Pipeline #12: Score on D2: 0.1390062362229828 | D1-D2 diff: 2.538277754989673 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=10, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1498615899811202 Holdout data R^2 trained on entire dataset(80%): 0.14839779614148008 Dataset D1 R^2 on trained D1: 0.16309660371915913 .................................................. Pipeline #13: Score on D2: 0.11126067669045259 | D1-D2 diff: 2.5851031328644094 Pipeline steps: SelectPercentile(percentile=35), HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=11, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13130809192964177 Holdout data R^2 trained on entire dataset(80%): 0.10077355027227186 Dataset D1 R^2 on trained D1: 0.1336524507893052 .................................................. Pipeline #14: Score on D2: 0.11008396266561404 | D1-D2 diff: 3.284113556977605 Pipeline steps: VarianceThreshold(threshold=0.25), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11561906228827445 Holdout data R^2 trained on entire dataset(80%): 0.13842516399868054 Dataset D1 R^2 on trained D1: 0.11868057490133366 .................................................. Pipeline #15: Score on D2: 0.10481085698648085 | D1-D2 diff: 4.020360061832426 Pipeline steps: SelectPercentile(percentile=90), DecisionTreeRegressor(max_depth=3, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10751623558838508 Holdout data R^2 trained on entire dataset(80%): 0.10576290599458638 Dataset D1 R^2 on trained D1: 0.10863857732466853 .................................................. Pipeline #16: Score on D2: 0.011960515030693841 | D1-D2 diff: 7.1550359484228325 Pipeline steps: DominantEncoder(), SelectPercentile(percentile=10), SelectPercentile(percentile=25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.012177163125235313 Holdout data R^2 trained on entire dataset(80%): 0.022271867387798294 Dataset D1 R^2 on trained D1: 0.012342066139210184 .................................................. Pipeline #17: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: UnderDominanceEncoder(), HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=4, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 32 - 7 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00000 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.00029 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00058 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.08730 Gen 2 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.89837 Gen 3 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 14 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 15 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 16 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 17 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 18 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 19 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 20 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 21 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 2 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 3 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.10063 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.10239 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 12 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.11435 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.12839 Best score on D1-D2 diff: 5.11886 Gen 3 - Best score on D2: 0.12839 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.12840 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.13233 Best score on D1-D2 diff: 16.97624 Gen 6 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 7 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 8 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 9 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 12 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 13 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 1 - Best score on D2: 0.14656 Best score on D1-D2 diff: 7.98775 Gen 2 - Best score on D2: 0.14969 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14969 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.16390 Best score on D1-D2 diff: 3.72276 Gen 2 - Best score on D2: 0.16463 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.16463 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.16463 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.16463 Best score on D1-D2 diff: 31.24909 Gen 6 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 7 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 8 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 9 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 10 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 11 - Best score on D2: 0.16823 Best score on D1-D2 diff: 31.24909 Gen 12 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 13 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 14 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 1 - Best score on D2: 0.16861 Best score on D1-D2 diff: 3.72276 Gen 2 - Best score on D2: 0.17263 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.17422 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.17422 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.17422 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.17422 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.17422 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.17422 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.18745 Best score on D1-D2 diff: 6.14285 Gen 2 - Best score on D2: 0.19248 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.19248 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.19624 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.20017 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.20017 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.011321838065091239 Entire dataset(80%) R^2 trained on data (80%): 0.0038438305342128887 Holdout R^2 (20%) trained on data (80%): 0.0008267139597373241 Dataset D1 R^2 on trained D1: 0.008840333510632958 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0038522111205001597 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.20017193400986377 | D1-D2 diff: 1.415374268977678 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4511339938629285 Holdout data R^2 trained on entire dataset(80%): 0.2334026066139181 Dataset D1 R^2 on trained D1: 0.44935287178819194 .................................................. Pipeline #2: Score on D2: 0.19927113783217276 | D1-D2 diff: 1.483232451142221 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=4, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.40938191259682566 Holdout data R^2 trained on entire dataset(80%): 0.2295964805328039 Dataset D1 R^2 on trained D1: 0.4058867456822277 .................................................. Pipeline #3: Score on D2: 0.1990340884517937 | D1-D2 diff: 1.5245512981185787 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3973645585196962 Holdout data R^2 trained on entire dataset(80%): 0.23220311789794545 Dataset D1 R^2 on trained D1: 0.38414491995262723 .................................................. Pipeline #4: Score on D2: 0.1976731394668878 | D1-D2 diff: 1.5419854787605476 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3861554632405275 Holdout data R^2 trained on entire dataset(80%): 0.22989666143844523 Dataset D1 R^2 on trained D1: 0.3745531944112356 .................................................. Pipeline #5: Score on D2: 0.19706007136551884 | D1-D2 diff: 1.5777063711534483 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(max_features=0.3, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3716494633687448 Holdout data R^2 trained on entire dataset(80%): 0.2335607126794731 Dataset D1 R^2 on trained D1: 0.3584570062532816 .................................................. Pipeline #6: Score on D2: 0.1958271944745814 | D1-D2 diff: 1.6145048735724863 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=2, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3570260176435216 Holdout data R^2 trained on entire dataset(80%): 0.22932427477136663 Dataset D1 R^2 on trained D1: 0.3430050798268329 .................................................. Pipeline #7: Score on D2: 0.19512296105623583 | D1-D2 diff: 1.6590632145482347 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=4, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3405663530294716 Holdout data R^2 trained on entire dataset(80%): 0.2300640534872056 Dataset D1 R^2 on trained D1: 0.32711515988042583 .................................................. Pipeline #8: Score on D2: 0.19218243147617053 | D1-D2 diff: 1.6740544966042672 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=11, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.32786000787705794 Holdout data R^2 trained on entire dataset(80%): 0.22268873020438407 Dataset D1 R^2 on trained D1: 0.319509761694571 .................................................. Pipeline #9: Score on D2: 0.1906058293059676 | D1-D2 diff: 1.7269481455591598 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=14, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3135508851510683 Holdout data R^2 trained on entire dataset(80%): 0.22151130554372733 Dataset D1 R^2 on trained D1: 0.30303598460533143 .................................................. Pipeline #10: Score on D2: 0.18824565253985093 | D1-D2 diff: 1.8378644634873864 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28806423212870313 Holdout data R^2 trained on entire dataset(80%): 0.21710208699607803 Dataset D1 R^2 on trained D1: 0.27589447347682594 .................................................. Pipeline #11: Score on D2: 0.18782957259671917 | D1-D2 diff: 1.8571999218440207 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=19, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2858080700594974 Holdout data R^2 trained on entire dataset(80%): 0.2172996720130692 Dataset D1 R^2 on trained D1: 0.2718849249705909 .................................................. Pipeline #12: Score on D2: 0.18749018099086134 | D1-D2 diff: 1.860952744569829 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=19, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2842382288751666 Holdout data R^2 trained on entire dataset(80%): 0.2161700183933395 Dataset D1 R^2 on trained D1: 0.27086955290437 .................................................. Pipeline #13: Score on D2: 0.18519653737092456 | D1-D2 diff: 1.8703058596265556 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=19, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2831833961901383 Holdout data R^2 trained on entire dataset(80%): 0.2142992919421005 Dataset D1 R^2 on trained D1: 0.26692050855887683 .................................................. Pipeline #14: Score on D2: 0.18046309316193077 | D1-D2 diff: 1.972606529124865 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), OverDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=15, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26418020638495066 Holdout data R^2 trained on entire dataset(80%): 0.21181769342054335 Dataset D1 R^2 on trained D1: 0.2465078179757415 .................................................. Pipeline #15: Score on D2: 0.17827820493945223 | D1-D2 diff: 2.0000195536161622 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2582754042176387 Holdout data R^2 trained on entire dataset(80%): 0.20949063477720464 Dataset D1 R^2 on trained D1: 0.240775760797172 .................................................. Pipeline #16: Score on D2: 0.1775415649093507 | D1-D2 diff: 2.014827964687734 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2575975444849211 Holdout data R^2 trained on entire dataset(80%): 0.20625744198682283 Dataset D1 R^2 on trained D1: 0.23822192090729544 .................................................. Pipeline #17: Score on D2: 0.17739757309457294 | D1-D2 diff: 2.0428711513406226 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.254467292462497 Holdout data R^2 trained on entire dataset(80%): 0.20914264905990054 Dataset D1 R^2 on trained D1: 0.2348139914811953 .................................................. Pipeline #18: Score on D2: 0.1766900225143072 | D1-D2 diff: 2.051261734656251 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25142353293724107 Holdout data R^2 trained on entire dataset(80%): 0.20403986881408087 Dataset D1 R^2 on trained D1: 0.23317275334521304 .................................................. Pipeline #19: Score on D2: 0.17633399458208454 | D1-D2 diff: 2.101235465935513 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=19, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24731706427638356 Holdout data R^2 trained on entire dataset(80%): 0.20500324147589466 Dataset D1 R^2 on trained D1: 0.22763207453093848 .................................................. Pipeline #20: Score on D2: 0.16811272066829397 | D1-D2 diff: 2.112387916938042 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23998948047759727 Holdout data R^2 trained on entire dataset(80%): 0.20054046940507342 Dataset D1 R^2 on trained D1: 0.21833602716999867 .................................................. Pipeline #21: Score on D2: 0.16645567087267665 | D1-D2 diff: 2.1333632781752576 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RecessiveEncoder(), OverDominanceEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23686255746083906 Holdout data R^2 trained on entire dataset(80%): 0.20262848130206024 Dataset D1 R^2 on trained D1: 0.2147327224923974 .................................................. Pipeline #22: Score on D2: 0.16596998441267907 | D1-D2 diff: 2.1618511619565086 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RecessiveEncoder(), OverDominanceEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=19, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23148987015855804 Holdout data R^2 trained on entire dataset(80%): 0.19988586259586427 Dataset D1 R^2 on trained D1: 0.21175220333616274 .................................................. Pipeline #23: Score on D2: 0.1636207625624967 | D1-D2 diff: 2.162721149163012 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RecessiveEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=19, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.232833510310557 Holdout data R^2 trained on entire dataset(80%): 0.19408285931637836 Dataset D1 R^2 on trained D1: 0.20932935957145005 .................................................. Pipeline #24: Score on D2: 0.1580226384406226 | D1-D2 diff: 2.3032138786736063 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RecessiveEncoder(), SelectPercentile(percentile=65), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18534753908368473 Holdout data R^2 trained on entire dataset(80%): 0.15213357880216116 Dataset D1 R^2 on trained D1: 0.19355817886896942 .................................................. Pipeline #25: Score on D2: 0.15536481682769687 | D1-D2 diff: 2.310090184454974 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RecessiveEncoder(), SelectPercentile(percentile=65), OverDominanceEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1858003397177277 Holdout data R^2 trained on entire dataset(80%): 0.15349392799262185 Dataset D1 R^2 on trained D1: 0.19047913668223615 .................................................. Pipeline #26: Score on D2: 0.1473933922122883 | D1-D2 diff: 2.340520654197332 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), UnderDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=17, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20821317843248832 Holdout data R^2 trained on entire dataset(80%): 0.17613101010159726 Dataset D1 R^2 on trained D1: 0.1807168522084256 .................................................. Pipeline #27: Score on D2: 0.12726478428777732 | D1-D2 diff: 2.416848097230835 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), UnderDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1773424067237097 Holdout data R^2 trained on entire dataset(80%): 0.15609827830276313 Dataset D1 R^2 on trained D1: 0.156573890811618 .................................................. Pipeline #28: Score on D2: 0.1259577889541913 | D1-D2 diff: 2.4283002326418366 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18194658101862404 Holdout data R^2 trained on entire dataset(80%): 0.156723119047946 Dataset D1 R^2 on trained D1: 0.15471789440914163 .................................................. Pipeline #29: Score on D2: 0.1241085777130253 | D1-D2 diff: 4.68303574129622 Pipeline steps: SelectPercentile(percentile=40), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), DecisionTreeRegressor(max_depth=4, min_samples_leaf=10, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05668405695956891 Holdout data R^2 trained on entire dataset(80%): 0.048749285976880774 Dataset D1 R^2 on trained D1: 0.12202940718464261 .................................................. Pipeline #30: Score on D2: 0.0660752431171352 | D1-D2 diff: 7.587922944151055 Pipeline steps: UnderDominanceEncoder(), RecessiveEncoder(), SelectPercentile(percentile=40), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0682824775641585 Holdout data R^2 trained on entire dataset(80%): 0.0658288174767594 Dataset D1 R^2 on trained D1: 0.06577358962858126 .................................................. Pipeline #31: Score on D2: 0.06569504159679063 | D1-D2 diff: 10.09291117452603 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=40), UnderDominanceEncoder(), DominantEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0682824775641585 Holdout data R^2 trained on entire dataset(80%): 0.0658288174767594 Dataset D1 R^2 on trained D1: 0.0655986732973175 .................................................. Pipeline #32: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: DominantEncoder(), OverDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=9, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 32 - 8 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00000 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.00029 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00058 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00127 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00322 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.08730 Gen 2 - Best score on D2: 0.09117 Best score on D1-D2 diff: 4.89837 Gen 3 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.09117 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.09117 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 14 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 15 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 16 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 17 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 18 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 19 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 20 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 21 - Best score on D2: 0.09117 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 2 - Best score on D2: 0.10044 Best score on D1-D2 diff: 3.74458 Gen 3 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.10063 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.10063 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.10239 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.10514 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 12 - Best score on D2: 0.10514 Best score on D1-D2 diff: 16.97624 Gen 1 - Best score on D2: 0.11435 Best score on D1-D2 diff: 4.89837 Gen 2 - Best score on D2: 0.12839 Best score on D1-D2 diff: 5.11886 Gen 3 - Best score on D2: 0.12839 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.12840 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.13233 Best score on D1-D2 diff: 16.97624 Gen 6 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 7 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 8 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 9 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 10 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97624 Gen 11 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 12 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 13 - Best score on D2: 0.13821 Best score on D1-D2 diff: 16.97962 Gen 1 - Best score on D2: 0.14656 Best score on D1-D2 diff: 7.98775 Gen 2 - Best score on D2: 0.14969 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14969 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15241 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.16390 Best score on D1-D2 diff: 3.72276 Gen 2 - Best score on D2: 0.16463 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.16463 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.16463 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.16463 Best score on D1-D2 diff: 31.24909 Gen 6 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 7 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 8 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 9 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 10 - Best score on D2: 0.16768 Best score on D1-D2 diff: 31.24909 Gen 11 - Best score on D2: 0.16823 Best score on D1-D2 diff: 31.24909 Gen 12 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 13 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 14 - Best score on D2: 0.16944 Best score on D1-D2 diff: 31.24909 Gen 1 - Best score on D2: 0.16861 Best score on D1-D2 diff: 3.72276 Gen 2 - Best score on D2: 0.17263 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.17422 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.17422 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.17422 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.17422 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.17422 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.17422 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.18745 Best score on D1-D2 diff: 6.14285 Gen 2 - Best score on D2: 0.19248 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.19248 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.19248 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.19624 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.20001 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.20017 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.20017 Best score on D1-D2 diff: 13.48015 Gen 1 - Best score on D2: 0.18287 Best score on D1-D2 diff: 11.72506 Gen 2 - Best score on D2: 0.18597 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.19118 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.19118 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.19118 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.19118 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.19118 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.19118 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.19339 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.19339 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.19339 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.19339 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.19339 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.19339 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.19339 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.19367 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.19367 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.19367 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.19367 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.19367 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.20156 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.20156 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.20156 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.20156 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.20156 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.012004939330892306 Entire dataset(80%) R^2 trained on data (80%): 0.0037848268739231195 Holdout R^2 (20%) trained on data (80%): 0.0010151474921279435 Dataset D1 R^2 on trained D1: 0.009353574491986705 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004164886504674836 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.20156185582670927 | D1-D2 diff: 1.4417853982913287 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.45, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.43910842331865196 Holdout data R^2 trained on entire dataset(80%): 0.25678831780159306 Dataset D1 R^2 on trained D1: 0.4329800543332748 .................................................. Pipeline #2: Score on D2: 0.20074341316924482 | D1-D2 diff: 1.4687557336603896 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4230874433133496 Holdout data R^2 trained on entire dataset(80%): 0.2603372528909077 Dataset D1 R^2 on trained D1: 0.4156262357485905 .................................................. Pipeline #3: Score on D2: 0.2000127094339189 | D1-D2 diff: 1.4760231316672314 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.4163776808428069 Holdout data R^2 trained on entire dataset(80%): 0.26265136427882385 Dataset D1 R^2 on trained D1: 0.4106946672290438 .................................................. Pipeline #4: Score on D2: 0.19884594208258444 | D1-D2 diff: 1.664474399288899 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.3, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.34147936437171633 Holdout data R^2 trained on entire dataset(80%): 0.2527635538241465 Dataset D1 R^2 on trained D1: 0.32913007314452447 .................................................. Pipeline #5: Score on D2: 0.19059992388669578 | D1-D2 diff: 1.7245428010544983 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=14, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3173857517630204 Holdout data R^2 trained on entire dataset(80%): 0.23919376265582504 Dataset D1 R^2 on trained D1: 0.30365865058838093 .................................................. Pipeline #6: Score on D2: 0.1873424734255278 | D1-D2 diff: 1.7970195215801634 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29548649670966753 Holdout data R^2 trained on entire dataset(80%): 0.23308608474592785 Dataset D1 R^2 on trained D1: 0.2832358961013045 .................................................. Pipeline #7: Score on D2: 0.18389265646494413 | D1-D2 diff: 1.8236543504127518 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=14, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2872091835984554 Holdout data R^2 trained on entire dataset(80%): 0.2309085071179836 Dataset D1 R^2 on trained D1: 0.2743054508041507 .................................................. Pipeline #8: Score on D2: 0.18192578243125246 | D1-D2 diff: 1.885656874103409 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=13, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.27645653647006796 Holdout data R^2 trained on entire dataset(80%): 0.2222997531650236 Dataset D1 R^2 on trained D1: 0.2610208361559434 .................................................. Pipeline #9: Score on D2: 0.17829590685424945 | D1-D2 diff: 2.0088813920225177 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2563580396414289 Holdout data R^2 trained on entire dataset(80%): 0.2226596131401507 Dataset D1 R^2 on trained D1: 0.23969794909730724 .................................................. Pipeline #10: Score on D2: 0.17462557378533694 | D1-D2 diff: 2.0213524603356126 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2554959200124629 Holdout data R^2 trained on entire dataset(80%): 0.2200051020945103 Dataset D1 R^2 on trained D1: 0.2345262617457956 .................................................. Pipeline #11: Score on D2: 0.17027585302948156 | D1-D2 diff: 2.0404154167784103 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=13, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2479029611712389 Holdout data R^2 trained on entire dataset(80%): 0.2125933669762955 Dataset D1 R^2 on trained D1: 0.2279691841179683 .................................................. Pipeline #12: Score on D2: 0.16903807124406667 | D1-D2 diff: 2.0483134921898123 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), RandomForestRegressor(max_features=0.4, min_samples_leaf=19, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2456701508073662 Holdout data R^2 trained on entire dataset(80%): 0.2144283456712751 Dataset D1 R^2 on trained D1: 0.22584669879514618 .................................................. Pipeline #13: Score on D2: 0.1672850136289341 | D1-D2 diff: 2.0899613486385493 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23975046284554857 Holdout data R^2 trained on entire dataset(80%): 0.20793509097520968 Dataset D1 R^2 on trained D1: 0.21969897476094424 .................................................. Pipeline #14: Score on D2: 0.16330056296875106 | D1-D2 diff: 2.141425904353409 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=75), RandomForestRegressor(max_features=0.4, min_samples_leaf=19, min_samples_split=11, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2260207959807905 Holdout data R^2 trained on entire dataset(80%): 0.19676579859889554 Dataset D1 R^2 on trained D1: 0.21085464386672415 .................................................. Pipeline #15: Score on D2: 0.14764391968472457 | D1-D2 diff: 2.2071158444139387 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=65), FeatureEncodingFrequencySelector(threshold=0.15), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21045832653113605 Holdout data R^2 trained on entire dataset(80%): 0.18882472877470713 Dataset D1 R^2 on trained D1: 0.18978440045679934 .................................................. Pipeline #16: Score on D2: 0.14387486305404062 | D1-D2 diff: 2.2611351563816795 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=65), OverDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21045832653113605 Holdout data R^2 trained on entire dataset(80%): 0.18882472877470713 Dataset D1 R^2 on trained D1: 0.18213036546893768 .................................................. Pipeline #17: Score on D2: 0.1426471104579612 | D1-D2 diff: 2.6230822586103657 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15448432572483106 Holdout data R^2 trained on entire dataset(80%): 0.1491053878844485 Dataset D1 R^2 on trained D1: 0.16376995290581442 .................................................. Pipeline #18: Score on D2: 0.11663365666415393 | D1-D2 diff: 2.7115182557647795 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.2), OverDominanceEncoder(), DominantEncoder(), SelectPercentile(percentile=95), DecisionTreeRegressor(max_depth=4, min_samples_leaf=15, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12171455847777357 Holdout data R^2 trained on entire dataset(80%): 0.1264229774340594 Dataset D1 R^2 on trained D1: 0.13513272552562094 .................................................. Pipeline #19: Score on D2: 0.11621329002278757 | D1-D2 diff: 3.050671314925042 Pipeline steps: VarianceThreshold(threshold=0.2), OverDominanceEncoder(), SelectPercentile(percentile=80), DecisionTreeRegressor(max_depth=4, min_samples_leaf=15, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13478895225928345 Holdout data R^2 trained on entire dataset(80%): 0.13744325195906526 Dataset D1 R^2 on trained D1: 0.12775893838000674 .................................................. Pipeline #20: Score on D2: 0.10486721668458787 | D1-D2 diff: 4.168099479283318 Pipeline steps: UnderDominanceEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10702610722169259 Holdout data R^2 trained on entire dataset(80%): 0.11307288364764967 Dataset D1 R^2 on trained D1: 0.1081804170258921 .................................................. Pipeline #21: Score on D2: 0.061181680170971475 | D1-D2 diff: 8.205299397928632 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0615143626295106 Holdout data R^2 trained on entire dataset(80%): 0.05654399437951341 Dataset D1 R^2 on trained D1: 0.061402288685605044 .................................................. Pipeline #22: Score on D2: 0.045013691194915384 | D1-D2 diff: 8.218247385334259 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), SelectPercentile(percentile=90), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04577955765017294 Holdout data R^2 trained on entire dataset(80%): 0.03301319889093435 Dataset D1 R^2 on trained D1: 0.045232912701989925 .................................................. Pipeline #23: Score on D2: 0.016039433611898746 | D1-D2 diff: 10.7576087901588 Pipeline steps: RecessiveEncoder(), SelectPercentile(percentile=5), RandomForestRegressor(max_features=0.7500000000000001, min_samples_leaf=7, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.016019057709417805 Holdout data R^2 trained on entire dataset(80%): 0.013035710880551332 Dataset D1 R^2 on trained D1: 0.015964765183061336 .................................................. Pipeline #24: Score on D2: 0.016022787543995 | D1-D2 diff: 11.725061285458576 Pipeline steps: VarianceThreshold(threshold=0.35), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01601948500429995 Holdout data R^2 trained on entire dataset(80%): 0.01307988257530135 Dataset D1 R^2 on trained D1: 0.01596987733044164 .................................................. Pipeline #25: Score on D2: 0.016022787543994776 | D1-D2 diff: 11.725061285470877 Pipeline steps: VarianceThreshold(threshold=0.2), DecisionTreeRegressor(max_depth=1, min_samples_leaf=3, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01601948500429995 Holdout data R^2 trained on entire dataset(80%): 0.01307988257530135 Dataset D1 R^2 on trained D1: 0.01596987733044164 .................................................. Pipeline #26: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=3, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 32 - 9 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.19202 Best score on D1-D2 diff: 6.12170 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.009654389567015143 Entire dataset(80%) R^2 trained on data (80%): 0.005124062521829176 Holdout R^2 (20%) trained on data (80%): 0.004650136895059331 Dataset D1 R^2 on trained D1: 0.011484513347465586 Combined Dataset (100%) R^2 trained on combined data (100%): 0.006067694873830853 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.19201928249601663 | D1-D2 diff: 1.787349989281372 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2997853794334956 Holdout data R^2 trained on entire dataset(80%): 0.2320859729304695 Dataset D1 R^2 on trained D1: 0.2900047325730246 .................................................. Pipeline #2: Score on D2: 0.14631765455120294 | D1-D2 diff: 1.7987343021324271 Pipeline steps: RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2619586389369798 Holdout data R^2 trained on entire dataset(80%): 0.1889905911491967 Dataset D1 R^2 on trained D1: 0.2418459289479168 .................................................. Pipeline #3: Score on D2: 0.1433257202067474 | D1-D2 diff: 1.8097974435693147 Pipeline steps: RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=19, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25781924192622185 Holdout data R^2 trained on entire dataset(80%): 0.1867382756750552 Dataset D1 R^2 on trained D1: 0.2365394998900059 .................................................. Pipeline #4: Score on D2: 0.1381737387588523 | D1-D2 diff: 1.9012133531243982 Pipeline steps: OverDominanceEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23402922911729052 Holdout data R^2 trained on entire dataset(80%): 0.17393881425792235 Dataset D1 R^2 on trained D1: 0.21471164480274463 .................................................. Pipeline #5: Score on D2: 0.1340930277749618 | D1-D2 diff: 1.9734567283789963 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=5, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17556580623652984 Holdout data R^2 trained on entire dataset(80%): 0.1439997123205191 Dataset D1 R^2 on trained D1: 0.2000240132824177 .................................................. Pipeline #6: Score on D2: 0.11060601227213462 | D1-D2 diff: 2.146339930644421 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15814027071703762 Holdout data R^2 trained on entire dataset(80%): 0.15454078838373897 Dataset D1 R^2 on trained D1: 0.15772608790311904 .................................................. Pipeline #7: Score on D2: 0.10755816537271512 | D1-D2 diff: 2.469900621629425 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.12012505879398205 Holdout data R^2 trained on entire dataset(80%): 0.12327972267931653 Dataset D1 R^2 on trained D1: 0.1344290573013066 .................................................. Pipeline #8: Score on D2: 0.08625335034823778 | D1-D2 diff: 2.982104907686707 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1030942967012849 Holdout data R^2 trained on entire dataset(80%): 0.11484603497447887 Dataset D1 R^2 on trained D1: 0.09889804452909823 .................................................. Pipeline #9: Score on D2: 0.07699835292403323 | D1-D2 diff: 3.62118068420194 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(threshold=0.25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0825524348208575 Holdout data R^2 trained on entire dataset(80%): 0.0686068761278491 Dataset D1 R^2 on trained D1: 0.08281401574935587 .................................................. Pipeline #10: Score on D2: 0.07239372949933998 | D1-D2 diff: 4.0026602322102685 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07460932525992181 Holdout data R^2 trained on entire dataset(80%): 0.04986779754174031 Dataset D1 R^2 on trained D1: 0.07628960522174455 .................................................. Pipeline #11: Score on D2: 0.03208176578103927 | D1-D2 diff: 6.121695843453064 Pipeline steps: RecessiveEncoder(), SelectPercentile(percentile=90), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03484841091644619 Holdout data R^2 trained on entire dataset(80%): 0.025995936941869968 Dataset D1 R^2 on trained D1: 0.03279381990903796 .................................................. ************************************************************************************** Random Seed 33 - 0 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: -0.00006 Best score on D1-D2 diff: 4.66962 Gen 2 - Best score on D2: 0.00013 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.00051 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.00051 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.00105 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.00105 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.00193 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.00193 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.00193 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.00193 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.00193 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.00193 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.00193 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.00193 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.00193 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.00221 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.00305 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.00305 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.00305 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.00305 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.00305 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.00349 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.00349 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.00349 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.00349 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.007629897518568685 Entire dataset(80%) R^2 trained on data (80%): 0.002017280442744962 Holdout R^2 (20%) trained on data (80%): -0.0005774300901391083 Dataset D1 R^2 on trained D1: 0.005147302398738285 Combined Dataset (100%) R^2 trained on combined data (100%): 0.002427156568102329 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.0034945502070381496 | D1-D2 diff: 2.161554768916174 Pipeline steps: RecessiveEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=5, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04610677290230236 Holdout data R^2 trained on entire dataset(80%): 0.0058128313156941 Dataset D1 R^2 on trained D1: 0.049301884982310695 .................................................. Pipeline #2: Score on D2: 0.0015929261740176281 | D1-D2 diff: 2.807715452589929 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=80), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=13, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011932740336374925 Holdout data R^2 trained on entire dataset(80%): 0.002506504863151404 Dataset D1 R^2 on trained D1: 0.017684096475314548 .................................................. Pipeline #3: Score on D2: 0.0009375802757370932 | D1-D2 diff: 3.153927133371773 Pipeline steps: RecessiveEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=18, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01342851283341906 Holdout data R^2 trained on entire dataset(80%): 0.0022633888868037744 Dataset D1 R^2 on trained D1: 0.0110439080259086 .................................................. Pipeline #4: Score on D2: 0.000818637922083032 | D1-D2 diff: 3.1699005580180315 Pipeline steps: RecessiveEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.05, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.011267576946065394 Holdout data R^2 trained on entire dataset(80%): 0.0017543479397388317 Dataset D1 R^2 on trained D1: 0.01072279333977133 .................................................. Pipeline #5: Score on D2: 0.0004899134504312563 | D1-D2 diff: 4.074654295769473 Pipeline steps: SelectPercentile(percentile=65), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0036356654942558686 Holdout data R^2 trained on entire dataset(80%): -0.0030426577495545892 Dataset D1 R^2 on trained D1: 0.004117659919880001 .................................................. Pipeline #6: Score on D2: 0.00041921184114268595 | D1-D2 diff: 7.6245860784967565 Pipeline steps: RecessiveEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=9, min_samples_split=10, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0006115464496325718 Holdout data R^2 trained on entire dataset(80%): -0.0001279360456940548 Dataset D1 R^2 on trained D1: 0.0007151049922102803 .................................................. Pipeline #7: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: DominantEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.8, min_samples_leaf=19, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 33 - 1 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.09112 Best score on D1-D2 diff: 3.62420 Gen 2 - Best score on D2: 0.10224 Best score on D1-D2 diff: 3.89378 Gen 3 - Best score on D2: 0.10224 Best score on D1-D2 diff: 3.89378 Gen 4 - Best score on D2: 0.10224 Best score on D1-D2 diff: 4.66962 Gen 5 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 6 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 7 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 8 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 9 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 10 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 11 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 12 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 13 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 14 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 15 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 16 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 17 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 18 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 19 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 20 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 21 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 22 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 23 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 24 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 Gen 25 - Best score on D2: 0.10224 Best score on D1-D2 diff: 24.34286 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.008180656403206088 Entire dataset(80%) R^2 trained on data (80%): 0.004081921495540319 Holdout R^2 (20%) trained on data (80%): 0.0006944432453988014 Dataset D1 R^2 on trained D1: 0.008544566457734826 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004279939105129693 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.10223577730366606 | D1-D2 diff: 3.1040178133523066 Pipeline steps: UnderDominanceEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=17, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10805513227576669 Holdout data R^2 trained on entire dataset(80%): 0.11690330659772019 Dataset D1 R^2 on trained D1: 0.11300794685050497 .................................................. Pipeline #2: Score on D2: 0.00033241942409434344 | D1-D2 diff: 24.34285678521109 Pipeline steps: SelectPercentile(percentile=5), DominantEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.000357410447612283 Holdout data R^2 trained on entire dataset(80%): 5.181172950874391e-05 Dataset D1 R^2 on trained D1: 0.0003352672525411826 .................................................. ************************************************************************************** Random Seed 33 - 2 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.09115 Best score on D1-D2 diff: 4.66962 Gen 2 - Best score on D2: 0.10469 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.10469 Best score on D1-D2 diff: 13.48015 Gen 4 - Best score on D2: 0.10469 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.010100318020123344 Entire dataset(80%) R^2 trained on data (80%): 0.0019082749539850452 Holdout R^2 (20%) trained on data (80%): 0.00016971132637255693 Dataset D1 R^2 on trained D1: 0.005558026057050913 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0023065321277224715 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.10469005449359747 | D1-D2 diff: 1.5262020148419153 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6500000000000001, min_samples_leaf=9, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.30069667653541954 Holdout data R^2 trained on entire dataset(80%): 0.13340992417532493 Dataset D1 R^2 on trained D1: 0.28900133218149193 .................................................. Pipeline #2: Score on D2: 0.10258621951764901 | D1-D2 diff: 1.783925147370624 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2234521648080251 Holdout data R^2 trained on entire dataset(80%): 0.13002374767695934 Dataset D1 R^2 on trained D1: 0.20132630285905517 .................................................. Pipeline #3: Score on D2: 0.09939584530775669 | D1-D2 diff: 1.8321390780704174 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=19, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2137257068335945 Holdout data R^2 trained on entire dataset(80%): 0.1283465812101735 Dataset D1 R^2 on trained D1: 0.1881454133834617 .................................................. Pipeline #4: Score on D2: 0.08975986435185823 | D1-D2 diff: 1.879587129478566 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19212138271857937 Holdout data R^2 trained on entire dataset(80%): 0.1254861181870548 Dataset D1 R^2 on trained D1: 0.1698815633311368 .................................................. Pipeline #5: Score on D2: 0.08689730506934235 | D1-D2 diff: 2.557144487827709 Pipeline steps: SelectPercentile(percentile=50), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.005984759802912132 Holdout data R^2 trained on entire dataset(80%): -0.004865755868968646 Dataset D1 R^2 on trained D1: 0.1102845426963962 .................................................. Pipeline #6: Score on D2: 0.008936711504062611 | D1-D2 diff: 2.6511499338552804 Pipeline steps: DominantEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.00867437307449792 Holdout data R^2 trained on entire dataset(80%): -0.003439220044681779 Dataset D1 R^2 on trained D1: 0.02917915060844345 .................................................. Pipeline #7: Score on D2: 0.0011857267289515638 | D1-D2 diff: 6.573796296326289 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=50), FeatureEncodingFrequencySelector(threshold=0.2), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0014720903752398762 Holdout data R^2 trained on entire dataset(80%): -0.000262309558006546 Dataset D1 R^2 on trained D1: 0.0006502568692193789 .................................................. Pipeline #8: Score on D2: 0.0005981212128934876 | D1-D2 diff: 8.789375260869694 Pipeline steps: UnderDominanceEncoder(), UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0033673988570850355 Holdout data R^2 trained on entire dataset(80%): 0.0007324528942367348 Dataset D1 R^2 on trained D1: 0.00043056213101999585 .................................................. Pipeline #9: Score on D2: 0.00012472021329457128 | D1-D2 diff: 9.555059116737084 Pipeline steps: UnderDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002735773513237283 Holdout data R^2 trained on entire dataset(80%): -0.0005395652091022551 Dataset D1 R^2 on trained D1: 4.751918324275017e-06 .................................................. Pipeline #10: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.8500000000000001, min_samples_leaf=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 33 - 3 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.12536 Best score on D1-D2 diff: 4.56074 Gen 2 - Best score on D2: 0.13015 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.13519 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.13519 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.13519 Best score on D1-D2 diff: 11.96036 Gen 6 - Best score on D2: 0.13519 Best score on D1-D2 diff: 11.96036 Gen 7 - Best score on D2: 0.13519 Best score on D1-D2 diff: 11.96036 Gen 8 - Best score on D2: 0.13851 Best score on D1-D2 diff: 11.96036 Gen 9 - Best score on D2: 0.13851 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.00790763014359297 Entire dataset(80%) R^2 trained on data (80%): 0.002181418554203729 Holdout R^2 (20%) trained on data (80%): 0.001045554830217621 Dataset D1 R^2 on trained D1: 0.005477716417912615 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0027483458833619245 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1385118570180598 | D1-D2 diff: 1.7823345008112992 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.8, min_samples_leaf=19, min_samples_split=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2445669004124169 Holdout data R^2 trained on entire dataset(80%): 0.16472785712969606 Dataset D1 R^2 on trained D1: 0.237604895323263 .................................................. Pipeline #2: Score on D2: 0.13430662374954605 | D1-D2 diff: 1.81797727495325 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.3), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2330778636539036 Holdout data R^2 trained on entire dataset(80%): 0.15390862560359853 Dataset D1 R^2 on trained D1: 0.22585406269754016 .................................................. Pipeline #3: Score on D2: 0.11390121782778029 | D1-D2 diff: 1.839715263321368 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.0), SelectPercentile(percentile=90), RandomForestRegressor(bootstrap=False, max_features=0.2, min_samples_leaf=18, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.21124582932950609 Holdout data R^2 trained on entire dataset(80%): 0.14188108791275178 Dataset D1 R^2 on trained D1: 0.2011978629821911 .................................................. Pipeline #4: Score on D2: 0.10698859843567188 | D1-D2 diff: 1.8424948552965457 Pipeline steps: SelectPercentile(percentile=65), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.8, min_samples_leaf=18, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19545288546088413 Holdout data R^2 trained on entire dataset(80%): 0.13638766664479018 Dataset D1 R^2 on trained D1: 0.19375965077383073 .................................................. Pipeline #5: Score on D2: 0.10369005385780583 | D1-D2 diff: 1.9141052350226826 Pipeline steps: SelectPercentile(percentile=65), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.5, min_samples_leaf=18, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18639301110497253 Holdout data R^2 trained on entire dataset(80%): 0.1427552728412831 Dataset D1 R^2 on trained D1: 0.17818670594901231 .................................................. Pipeline #6: Score on D2: 0.1025293616354106 | D1-D2 diff: 3.1162649794861346 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.09782376611986099 Holdout data R^2 trained on entire dataset(80%): 0.11039440252884836 Dataset D1 R^2 on trained D1: 0.11313318496999614 .................................................. Pipeline #7: Score on D2: 0.018353270828943336 | D1-D2 diff: 3.744606847785666 Pipeline steps: VarianceThreshold(threshold=0.3), UnderDominanceEncoder(), HeterosisEncoder(), RecessiveEncoder(), RecessiveEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03815866763698694 Holdout data R^2 trained on entire dataset(80%): 0.045531387077282326 Dataset D1 R^2 on trained D1: 0.023439256020603705 .................................................. Pipeline #8: Score on D2: 0.005791483672755282 | D1-D2 diff: 4.761643913704678 Pipeline steps: RecessiveEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.007357641309568397 Holdout data R^2 trained on entire dataset(80%): 0.002603882163620841 Dataset D1 R^2 on trained D1: 0.007736719863276398 .................................................. Pipeline #9: Score on D2: 0.0007066893712094346 | D1-D2 diff: 9.057609629521302 Pipeline steps: SelectPercentile(percentile=15), SelectPercentile(percentile=55), RecessiveEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0008304363594029418 Holdout data R^2 trained on entire dataset(80%): -0.0009055613472919166 Dataset D1 R^2 on trained D1: 0.0008552643246968472 .................................................. Pipeline #10: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: SelectPercentile(percentile=10), RecessiveEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=5, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0003740462455762428 Holdout data R^2 trained on entire dataset(80%): -0.000721821267699152 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 33 - 4 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.14381 Best score on D1-D2 diff: 4.67432 Gen 2 - Best score on D2: 0.14970 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.14970 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15403 Best score on D1-D2 diff: 13.48015 Gen 5 - Best score on D2: 0.15678 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.15678 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.15748 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.15748 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.15748 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.15748 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.15748 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.15748 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.16006 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.16006 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.16006 Best score on D1-D2 diff: 13.48015 Gen 16 - Best score on D2: 0.16006 Best score on D1-D2 diff: 13.48015 Gen 17 - Best score on D2: 0.16006 Best score on D1-D2 diff: 13.48015 Gen 18 - Best score on D2: 0.16282 Best score on D1-D2 diff: 13.48015 Gen 19 - Best score on D2: 0.16282 Best score on D1-D2 diff: 13.48015 Gen 20 - Best score on D2: 0.16282 Best score on D1-D2 diff: 13.48015 Gen 21 - Best score on D2: 0.16282 Best score on D1-D2 diff: 13.48015 Gen 22 - Best score on D2: 0.16282 Best score on D1-D2 diff: 13.48015 Gen 23 - Best score on D2: 0.16573 Best score on D1-D2 diff: 13.48015 Gen 24 - Best score on D2: 0.16573 Best score on D1-D2 diff: 13.48015 Gen 25 - Best score on D2: 0.16573 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.007591493425660234 Entire dataset(80%) R^2 trained on data (80%): 0.0028189836302728866 Holdout R^2 (20%) trained on data (80%): -0.0007841595366639975 Dataset D1 R^2 on trained D1: 0.0062499347093365465 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0028281503087443927 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.165729648874972 | D1-D2 diff: 1.6371125925793255 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), SelectPercentile(percentile=80), RandomForestRegressor(max_features=0.5, min_samples_leaf=4, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.31335017689614086 Holdout data R^2 trained on entire dataset(80%): 0.1880215282846629 Dataset D1 R^2 on trained D1: 0.3049445757672644 .................................................. Pipeline #2: Score on D2: 0.1654232943885181 | D1-D2 diff: 1.8879183294083162 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), SelectPercentile(percentile=80), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=13, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.25592855764053934 Holdout data R^2 trained on entire dataset(80%): 0.18230176710660873 Dataset D1 R^2 on trained D1: 0.24414005040432285 .................................................. Pipeline #3: Score on D2: 0.16164524070435038 | D1-D2 diff: 1.9752140603544062 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), SelectPercentile(percentile=80), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=13, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24068775514573337 Holdout data R^2 trained on entire dataset(80%): 0.17974276410559265 Dataset D1 R^2 on trained D1: 0.22734190609440774 .................................................. Pipeline #4: Score on D2: 0.15581201018385338 | D1-D2 diff: 1.9929290317196982 Pipeline steps: HeterosisEncoder(), SelectPercentile(percentile=80), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=13, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2332048234341555 Holdout data R^2 trained on entire dataset(80%): 0.1694941233182402 Dataset D1 R^2 on trained D1: 0.21920374908306273 .................................................. Pipeline #5: Score on D2: 0.15127892530466447 | D1-D2 diff: 2.0112024760078357 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.25), SelectPercentile(percentile=80), RandomForestRegressor(max_features=0.25, min_samples_leaf=13, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.228734313856096 Holdout data R^2 trained on entire dataset(80%): 0.17264045096342717 Dataset D1 R^2 on trained D1: 0.21239800693911182 .................................................. Pipeline #6: Score on D2: 0.15078299243698545 | D1-D2 diff: 2.136170935368026 Pipeline steps: SelectPercentile(percentile=65), HeterosisEncoder(), VarianceThreshold(threshold=0.2), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=20, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.18119964155590296 Holdout data R^2 trained on entire dataset(80%): 0.13518936541314241 Dataset D1 R^2 on trained D1: 0.19880673396113002 .................................................. Pipeline #7: Score on D2: 0.14695383367796955 | D1-D2 diff: 3.2586043141945567 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=30), FeatureEncodingFrequencySelector(threshold=0.05), DominantEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11500583405714493 Holdout data R^2 trained on entire dataset(80%): 0.09816762577142812 Dataset D1 R^2 on trained D1: 0.15582280983529107 .................................................. Pipeline #8: Score on D2: 0.11298302977377128 | D1-D2 diff: 3.517686421769182 Pipeline steps: SelectPercentile(percentile=45), VarianceThreshold(threshold=0.3), HeterosisEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=17, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.08718706387139152 Holdout data R^2 trained on entire dataset(80%): 0.08382567302407207 Dataset D1 R^2 on trained D1: 0.10645215268492203 .................................................. Pipeline #9: Score on D2: 0.1041877530132963 | D1-D2 diff: 3.5307576739523516 Pipeline steps: SelectPercentile(percentile=45), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07007825736805573 Holdout data R^2 trained on entire dataset(80%): 0.06463806151410467 Dataset D1 R^2 on trained D1: 0.09775305228386422 .................................................. Pipeline #10: Score on D2: 0.05932340311203099 | D1-D2 diff: 5.346765845728454 Pipeline steps: SelectPercentile(percentile=45), HeterosisEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.006021533111840172 Holdout data R^2 trained on entire dataset(80%): -0.006610499526410685 Dataset D1 R^2 on trained D1: 0.058099814745358414 .................................................. Pipeline #11: Score on D2: 0.021526560894189695 | D1-D2 diff: 5.422281417997264 Pipeline steps: SelectPercentile(percentile=70), SelectPercentile(percentile=95), DecisionTreeRegressor(max_depth=3, min_samples_leaf=13, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.07591851790074178 Holdout data R^2 trained on entire dataset(80%): 0.07931746574170218 Dataset D1 R^2 on trained D1: 0.022683396866245764 .................................................. Pipeline #12: Score on D2: 0.01603884378472009 | D1-D2 diff: 5.974948963692669 Pipeline steps: SelectPercentile(percentile=40), VarianceThreshold(threshold=0.35), DecisionTreeRegressor(max_depth=2, min_samples_leaf=17, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0024305238478614655 Holdout data R^2 trained on entire dataset(80%): -0.0019921497245130038 Dataset D1 R^2 on trained D1: 0.015254216872768844 .................................................. Pipeline #13: Score on D2: 0.00041921184114268595 | D1-D2 diff: 7.6245860784967565 Pipeline steps: RecessiveEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=9, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0009094258717118331 Holdout data R^2 trained on entire dataset(80%): 0.0012919080584687936 Dataset D1 R^2 on trained D1: 0.0007151049922102803 .................................................. Pipeline #14: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: HeterosisEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=13, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 33 - 5 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.15570 Best score on D1-D2 diff: 5.00879 Gen 2 - Best score on D2: 0.15570 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.15793 Best score on D1-D2 diff: 11.96036 Gen 4 - Best score on D2: 0.15793 Best score on D1-D2 diff: 11.96036 Gen 5 - Best score on D2: 0.16403 Best score on D1-D2 diff: 13.48015 Gen 6 - Best score on D2: 0.16403 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.16403 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.16403 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.16403 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.16403 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.16403 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.16403 Best score on D1-D2 diff: 14.63925 Gen 13 - Best score on D2: 0.16403 Best score on D1-D2 diff: 14.63925 Gen 14 - Best score on D2: 0.16403 Best score on D1-D2 diff: 14.63925 Gen 15 - Best score on D2: 0.16403 Best score on D1-D2 diff: 14.63925 Gen 16 - Best score on D2: 0.16403 Best score on D1-D2 diff: 14.63925 Gen 17 - Best score on D2: 0.16403 Best score on D1-D2 diff: 14.63925 Gen 18 - Best score on D2: 0.16403 Best score on D1-D2 diff: 14.63925 Gen 19 - Best score on D2: 0.16403 Best score on D1-D2 diff: 14.63925 Gen 20 - Best score on D2: 0.16480 Best score on D1-D2 diff: 14.63925 Gen 21 - Best score on D2: 0.16480 Best score on D1-D2 diff: 14.63925 Gen 22 - Best score on D2: 0.16480 Best score on D1-D2 diff: 14.63925 Gen 23 - Best score on D2: 0.16480 Best score on D1-D2 diff: 14.63925 Gen 24 - Best score on D2: 0.16480 Best score on D1-D2 diff: 14.63925 Gen 25 - Best score on D2: 0.16480 Best score on D1-D2 diff: 14.63925 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0034878065142780468 Entire dataset(80%) R^2 trained on data (80%): 0.003190081723213112 Holdout R^2 (20%) trained on data (80%): -4.7362336740519595e-05 Dataset D1 R^2 on trained D1: 0.004533809684392431 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0032228531631759427 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1647961691726726 | D1-D2 diff: 1.673143621313527 Pipeline steps: OverDominanceEncoder(), VarianceThreshold(threshold=0.25), DominantEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.55, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3015820969883293 Holdout data R^2 trained on entire dataset(80%): 0.18001717414739304 Dataset D1 R^2 on trained D1: 0.29240099873350245 .................................................. Pipeline #2: Score on D2: 0.16286707254163968 | D1-D2 diff: 1.705412431794945 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.45, min_samples_leaf=19, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2912080823139438 Holdout data R^2 trained on entire dataset(80%): 0.1805202170244704 Dataset D1 R^2 on trained D1: 0.28108471711771 .................................................. Pipeline #3: Score on D2: 0.16003912353748606 | D1-D2 diff: 1.7265754355659395 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28132360286552693 Holdout data R^2 trained on entire dataset(80%): 0.18362326234460746 Dataset D1 R^2 on trained D1: 0.2725663899435402 .................................................. Pipeline #4: Score on D2: 0.15970929188791316 | D1-D2 diff: 1.7497797529979362 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.28026423665745814 Holdout data R^2 trained on entire dataset(80%): 0.18391306555438602 Dataset D1 R^2 on trained D1: 0.26638522546543175 .................................................. Pipeline #5: Score on D2: 0.15675461848349048 | D1-D2 diff: 1.778655280626852 Pipeline steps: HeterosisEncoder(), VarianceThreshold(threshold=0.15), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26970051672470374 Holdout data R^2 trained on entire dataset(80%): 0.18552996543239142 Dataset D1 R^2 on trained D1: 0.2566701161150521 .................................................. Pipeline #6: Score on D2: 0.15632631342400993 | D1-D2 diff: 1.8086876672400949 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26332731292084677 Holdout data R^2 trained on entire dataset(80%): 0.1792199110967153 Dataset D1 R^2 on trained D1: 0.2497690805576338 .................................................. Pipeline #7: Score on D2: 0.15529392577689705 | D1-D2 diff: 1.8293633939199014 Pipeline steps: OverDominanceEncoder(), VarianceThreshold(threshold=0.25), DominantEncoder(), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2579924086642248 Holdout data R^2 trained on entire dataset(80%): 0.18135728593364164 Dataset D1 R^2 on trained D1: 0.24458335814659982 .................................................. Pipeline #8: Score on D2: 0.15483513121081072 | D1-D2 diff: 1.8927263219890273 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=19, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2490352388414082 Holdout data R^2 trained on entire dataset(80%): 0.17963017604724474 Dataset D1 R^2 on trained D1: 0.23275508970592684 .................................................. Pipeline #9: Score on D2: 0.15261420632644473 | D1-D2 diff: 1.8966715204139244 Pipeline steps: VarianceThreshold(threshold=0.15), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=19, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24882656708818307 Holdout data R^2 trained on entire dataset(80%): 0.18185438859086034 Dataset D1 R^2 on trained D1: 0.2298878707727332 .................................................. Pipeline #10: Score on D2: 0.15164283080528929 | D1-D2 diff: 1.9534348255517846 Pipeline steps: VarianceThreshold(threshold=0.1), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23723286931527265 Holdout data R^2 trained on entire dataset(80%): 0.1756991330787071 Dataset D1 R^2 on trained D1: 0.2203187207581766 .................................................. Pipeline #11: Score on D2: 0.1515674970726243 | D1-D2 diff: 2.024502710466007 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=16, min_samples_split=19, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.22371799892192734 Holdout data R^2 trained on entire dataset(80%): 0.16887953387321608 Dataset D1 R^2 on trained D1: 0.2110962178182727 .................................................. Pipeline #12: Score on D2: 0.1279716401890949 | D1-D2 diff: 2.0746632134928418 Pipeline steps: VarianceThreshold(threshold=0.1), SelectPercentile(percentile=70), HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=19, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.17405931603348634 Holdout data R^2 trained on entire dataset(80%): 0.14438355885246135 Dataset D1 R^2 on trained D1: 0.18194874350331836 .................................................. Pipeline #13: Score on D2: 0.12314592796375734 | D1-D2 diff: 2.124230022634932 Pipeline steps: VarianceThreshold(threshold=0.1), HeterosisEncoder(), DecisionTreeRegressor(max_depth=5, min_samples_leaf=17, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15158032756758255 Holdout data R^2 trained on entire dataset(80%): 0.12611541480815924 Dataset D1 R^2 on trained D1: 0.1722586301307275 .................................................. Pipeline #14: Score on D2: 0.12067523887382814 | D1-D2 diff: 2.4557201426584303 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=20, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.13752713040978748 Holdout data R^2 trained on entire dataset(80%): 0.127619167450014 Dataset D1 R^2 on trained D1: 0.14817218799662535 .................................................. Pipeline #15: Score on D2: 0.10739904539149792 | D1-D2 diff: 3.44702891949656 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=25), VarianceThreshold(threshold=0.25), DominantEncoder(), VarianceThreshold(threshold=0.2), RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=13, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.047047130198112685 Holdout data R^2 trained on entire dataset(80%): 0.042131016494251106 Dataset D1 R^2 on trained D1: 0.11448209513685415 .................................................. Pipeline #16: Score on D2: 0.05268979575240562 | D1-D2 diff: 6.106741290258276 Pipeline steps: DecisionTreeRegressor(max_depth=2, min_samples_leaf=15, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.052427650498206635 Holdout data R^2 trained on entire dataset(80%): 0.061008763575668645 Dataset D1 R^2 on trained D1: 0.05197074107889821 .................................................. Pipeline #17: Score on D2: 0.05268979575240551 | D1-D2 diff: 6.106741290258512 Pipeline steps: VarianceThreshold(threshold=0.15), DecisionTreeRegressor(max_depth=2, min_samples_leaf=15, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.052427650498206635 Holdout data R^2 trained on entire dataset(80%): 0.061008763575668534 Dataset D1 R^2 on trained D1: 0.05197074107889821 .................................................. Pipeline #18: Score on D2: 0.0526897957524054 | D1-D2 diff: 6.106741290258747 Pipeline steps: SelectPercentile(percentile=95), DecisionTreeRegressor(max_depth=2, min_samples_leaf=17, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05242765049820686 Holdout data R^2 trained on entire dataset(80%): 0.061008763575668534 Dataset D1 R^2 on trained D1: 0.05197074107889821 .................................................. Pipeline #19: Score on D2: 0.030636037353999557 | D1-D2 diff: 8.357205952793306 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), UnderDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03225281840672123 Holdout data R^2 trained on entire dataset(80%): 0.04122757505279118 Dataset D1 R^2 on trained D1: 0.030841038165510537 .................................................. Pipeline #20: Score on D2: 0.016109311584617925 | D1-D2 diff: 8.790496149085147 Pipeline steps: RecessiveEncoder(), SelectPercentile(percentile=10), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.01628140722487692 Holdout data R^2 trained on entire dataset(80%): 0.011563843434814758 Dataset D1 R^2 on trained D1: 0.016276785220083245 .................................................. Pipeline #21: Score on D2: 0.0036379985826469063 | D1-D2 diff: 13.760462440948306 Pipeline steps: SelectPercentile(percentile=10), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.005507543396963244 Holdout data R^2 trained on entire dataset(80%): 0.0054032472571556855 Dataset D1 R^2 on trained D1: 0.00366588982651439 .................................................. Pipeline #22: Score on D2: 0.0018794544870286423 | D1-D2 diff: 14.639245686181255 Pipeline steps: UnderDominanceEncoder(), OverDominanceEncoder(), SelectPercentile(percentile=10), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002189371383274241 Holdout data R^2 trained on entire dataset(80%): 0.001241323036387354 Dataset D1 R^2 on trained D1: 0.0019012278344358036 .................................................. ************************************************************************************** Random Seed 33 - 6 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.16468 Best score on D1-D2 diff: 5.62197 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0027637530930086918 Entire dataset(80%) R^2 trained on data (80%): 0.003092006927780222 Holdout R^2 (20%) trained on data (80%): -0.00012408520062723305 Dataset D1 R^2 on trained D1: 0.004120168712012373 Combined Dataset (100%) R^2 trained on combined data (100%): 0.0032266863301467774 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1646771918576584 | D1-D2 diff: 1.717345375444794 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2854190160785016 Holdout data R^2 trained on entire dataset(80%): 0.19862502124413162 Dataset D1 R^2 on trained D1: 0.27964319159011275 .................................................. Pipeline #2: Score on D2: 0.12200376691496984 | D1-D2 diff: 1.749994767317724 Pipeline steps: RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2367707133735919 Holdout data R^2 trained on entire dataset(80%): 0.16237102897640365 Dataset D1 R^2 on trained D1: 0.22862728290395185 .................................................. Pipeline #3: Score on D2: 0.1201322067106041 | D1-D2 diff: 1.9856649449367605 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=5, min_samples_leaf=3, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.16906628917969502 Holdout data R^2 trained on entire dataset(80%): 0.12918362832854846 Dataset D1 R^2 on trained D1: 0.18445666311635833 .................................................. Pipeline #4: Score on D2: 0.09675508614671202 | D1-D2 diff: 2.1370037187936957 Pipeline steps: DecisionTreeRegressor(max_depth=5, min_samples_leaf=20, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1303945192081628 Holdout data R^2 trained on entire dataset(80%): 0.10135460807218388 Dataset D1 R^2 on trained D1: 0.14470401263189747 .................................................. Pipeline #5: Score on D2: 0.09387068014402467 | D1-D2 diff: 2.5966159627226024 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10813366359102394 Holdout data R^2 trained on entire dataset(80%): 0.11775583877643692 Dataset D1 R^2 on trained D1: 0.115867966561499 .................................................. Pipeline #6: Score on D2: 0.09223758537882543 | D1-D2 diff: 2.802374740145783 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10079294512779824 Holdout data R^2 trained on entire dataset(80%): 0.10759549041497884 Dataset D1 R^2 on trained D1: 0.10845177176917631 .................................................. Pipeline #7: Score on D2: 0.0429764510983387 | D1-D2 diff: 3.2918614323491453 Pipeline steps: HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05232006586165794 Holdout data R^2 trained on entire dataset(80%): 0.06494730296282802 Dataset D1 R^2 on trained D1: 0.051492415102911915 .................................................. Pipeline #8: Score on D2: 0.042431499193168576 | D1-D2 diff: 4.115290197823062 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.04086401381016491 Holdout data R^2 trained on entire dataset(80%): 0.0515604158133538 Dataset D1 R^2 on trained D1: 0.038944931216443934 .................................................. Pipeline #9: Score on D2: 0.037029536195511636 | D1-D2 diff: 5.008785920388158 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=1, min_samples_leaf=12, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.03623829796794631 Holdout data R^2 trained on entire dataset(80%): 0.04368571936719323 Dataset D1 R^2 on trained D1: 0.03544073294356276 .................................................. Pipeline #10: Score on D2: 0.02017775671680122 | D1-D2 diff: 5.621974596837486 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.02169864296673385 Holdout data R^2 trained on entire dataset(80%): 0.020973013848587407 Dataset D1 R^2 on trained D1: 0.021178780703914657 .................................................. ************************************************************************************** Random Seed 33 - 7 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.17249 Best score on D1-D2 diff: 4.50708 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.0072921675374006956 Entire dataset(80%) R^2 trained on data (80%): 0.004617955141103081 Holdout R^2 (20%) trained on data (80%): -0.0028256235334820357 Dataset D1 R^2 on trained D1: 0.009031131098846013 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004652767553071646 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.17248959043894885 | D1-D2 diff: 1.6977730936198625 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29496984357057365 Holdout data R^2 trained on entire dataset(80%): 0.21708873626868663 Dataset D1 R^2 on trained D1: 0.29284937831679725 .................................................. Pipeline #2: Score on D2: 0.13232967378754845 | D1-D2 diff: 1.7234669608314093 Pipeline steps: RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2484615139742663 Holdout data R^2 trained on entire dataset(80%): 0.1704483422881875 Dataset D1 R^2 on trained D1: 0.24567096364023822 .................................................. Pipeline #3: Score on D2: 0.13224211922224682 | D1-D2 diff: 1.7496060703261471 Pipeline steps: RandomForestRegressor(max_features=0.55, min_samples_leaf=18, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24219376463079267 Holdout data R^2 trained on entire dataset(80%): 0.16710050325514847 Dataset D1 R^2 on trained D1: 0.23896041781100075 .................................................. Pipeline #4: Score on D2: 0.12810748680782735 | D1-D2 diff: 2.2689292058752764 Pipeline steps: OverDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.1544640121069094 Holdout data R^2 trained on entire dataset(80%): 0.15513174343010006 Dataset D1 R^2 on trained D1: 0.16584004220001602 .................................................. Pipeline #5: Score on D2: 0.09979439162412462 | D1-D2 diff: 2.6162104938663235 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11391711342574296 Holdout data R^2 trained on entire dataset(80%): 0.11874518811940715 Dataset D1 R^2 on trained D1: 0.12114003585318067 .................................................. Pipeline #6: Score on D2: 0.09106284825796207 | D1-D2 diff: 2.7476205429544915 Pipeline steps: UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=2, min_samples_leaf=5, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10067939507228552 Holdout data R^2 trained on entire dataset(80%): 0.10803441561909444 Dataset D1 R^2 on trained D1: 0.10860864043025953 .................................................. Pipeline #7: Score on D2: 0.052500534525344866 | D1-D2 diff: 2.89053935258877 Pipeline steps: HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06537711110913691 Holdout data R^2 trained on entire dataset(80%): 0.06632437263533142 Dataset D1 R^2 on trained D1: 0.0668251988767834 .................................................. Pipeline #8: Score on D2: 0.04900384754064635 | D1-D2 diff: 4.507082738365313 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=15), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.047941359780469206 Holdout data R^2 trained on entire dataset(80%): 0.051533315515975286 Dataset D1 R^2 on trained D1: 0.046580487863217 .................................................. ************************************************************************************** Random Seed 33 - 8 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.19420 Best score on D1-D2 diff: 8.83226 Gen 2 - Best score on D2: 0.19420 Best score on D1-D2 diff: 11.96036 Gen 3 - Best score on D2: 0.19420 Best score on D1-D2 diff: 13.48015 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.004050513896468599 Entire dataset(80%) R^2 trained on data (80%): 0.0050493984768607 Holdout R^2 (20%) trained on data (80%): -0.0002357870007949625 Dataset D1 R^2 on trained D1: 0.007673719900876397 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004814166820916621 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.1942010292603098 | D1-D2 diff: 1.7848034561839672 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3008874491648674 Holdout data R^2 trained on entire dataset(80%): 0.21132901751867672 Dataset D1 R^2 on trained D1: 0.2927468944874827 .................................................. Pipeline #2: Score on D2: 0.1896281170499714 | D1-D2 diff: 1.828862981514174 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.6000000000000001, min_samples_leaf=17, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2884867869259641 Holdout data R^2 trained on entire dataset(80%): 0.20776514245455602 Dataset D1 R^2 on trained D1: 0.2790153148226815 .................................................. Pipeline #3: Score on D2: 0.17596665422793678 | D1-D2 diff: 1.9589338835261907 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=12, min_samples_split=14, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26016976653834223 Holdout data R^2 trained on entire dataset(80%): 0.2005644445433522 Dataset D1 R^2 on trained D1: 0.24387464593768204 .................................................. Pipeline #4: Score on D2: 0.1537930865567967 | D1-D2 diff: 1.9932455909428395 Pipeline steps: OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.1, min_samples_leaf=8, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23771069170948256 Holdout data R^2 trained on entire dataset(80%): 0.17708323989088404 Dataset D1 R^2 on trained D1: 0.21714456456749387 .................................................. Pipeline #5: Score on D2: 0.13960929197800598 | D1-D2 diff: 2.4009109912915125 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=8, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15746467126537078 Holdout data R^2 trained on entire dataset(80%): 0.15250300563534436 Dataset D1 R^2 on trained D1: 0.16970438990236636 .................................................. Pipeline #6: Score on D2: 0.10006696033944773 | D1-D2 diff: 2.6881000986340235 Pipeline steps: HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.11144280591263278 Holdout data R^2 trained on entire dataset(80%): 0.10998182720614313 Dataset D1 R^2 on trained D1: 0.11921914201609918 .................................................. Pipeline #7: Score on D2: 0.0966334792887582 | D1-D2 diff: 3.7493690076363864 Pipeline steps: SelectPercentile(percentile=20), HeterosisEncoder(), DecisionTreeRegressor(max_depth=3, min_samples_split=7, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.0022176977694822186 Holdout data R^2 trained on entire dataset(80%): -2.7041016454232292e-05 Dataset D1 R^2 on trained D1: 0.10169367436010768 .................................................. Pipeline #8: Score on D2: 0.09649270256854725 | D1-D2 diff: 3.882928339944473 Pipeline steps: SelectPercentile(percentile=15), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.002196688680349501 Holdout data R^2 trained on entire dataset(80%): 0.0007323112630616135 Dataset D1 R^2 on trained D1: 0.10089178902544138 .................................................. Pipeline #9: Score on D2: 0.06587307726714697 | D1-D2 diff: 3.938093436778485 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), OverDominanceEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06757682783408814 Holdout data R^2 trained on entire dataset(80%): 0.06928337014407782 Dataset D1 R^2 on trained D1: 0.06171535059695665 .................................................. Pipeline #10: Score on D2: 0.06587307726714686 | D1-D2 diff: 3.938093436778511 Pipeline steps: HeterosisEncoder(), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.06757682783408814 Holdout data R^2 trained on entire dataset(80%): 0.06928337014407782 Dataset D1 R^2 on trained D1: 0.06171535059695665 .................................................. Pipeline #11: Score on D2: 0.05641463012288339 | D1-D2 diff: 5.576536143762978 Pipeline steps: OverDominanceEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.05898931980450339 Holdout data R^2 trained on entire dataset(80%): 0.06487416212329278 Dataset D1 R^2 on trained D1: 0.05538057922877038 .................................................. Pipeline #12: Score on D2: 0.01956414295327691 | D1-D2 diff: 8.832258807438583 Pipeline steps: RecessiveEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), LinearRegression() Entire dataset(80%) R^2 trained on entire dataset(80%): 0.021337214433279184 Holdout data R^2 trained on entire dataset(80%): 0.019004562312889606 Dataset D1 R^2 on trained D1: 0.01939981446726924 .................................................. Pipeline #13: Score on D2: -3.2015592299483586e-05 | D1-D2 diff: 13.480151843729264 Pipeline steps: RecessiveEncoder(), OverDominanceEncoder(), RecessiveEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=16, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): -8.051021425092841e-07 Holdout data R^2 trained on entire dataset(80%): -0.0012636516749739979 Dataset D1 R^2 on trained D1: -1.731060551568575e-06 .................................................. ************************************************************************************** Random Seed 33 - 9 Interactions ************************************************************************************** autoQTL using following parameters: population size = 100 offspring_size = None generations = 25 mutation rate = 0.9 crossover rate = 0.1 ------------------------------------------------- Evolution History: Gen 1 - Best score on D2: 0.20409 Best score on D1-D2 diff: 11.32479 Gen 2 - Best score on D2: 0.20409 Best score on D1-D2 diff: 11.32479 Gen 3 - Best score on D2: 0.20547 Best score on D1-D2 diff: 12.80981 Gen 4 - Best score on D2: 0.20547 Best score on D1-D2 diff: 12.80981 Gen 5 - Best score on D2: 0.20547 Best score on D1-D2 diff: 12.80981 Gen 6 - Best score on D2: 0.20723 Best score on D1-D2 diff: 13.48015 Gen 7 - Best score on D2: 0.21716 Best score on D1-D2 diff: 13.48015 Gen 8 - Best score on D2: 0.21716 Best score on D1-D2 diff: 13.48015 Gen 9 - Best score on D2: 0.21716 Best score on D1-D2 diff: 13.48015 Gen 10 - Best score on D2: 0.21716 Best score on D1-D2 diff: 13.48015 Gen 11 - Best score on D2: 0.21716 Best score on D1-D2 diff: 13.48015 Gen 12 - Best score on D2: 0.21716 Best score on D1-D2 diff: 13.48015 Gen 13 - Best score on D2: 0.21716 Best score on D1-D2 diff: 13.48015 Gen 14 - Best score on D2: 0.21716 Best score on D1-D2 diff: 13.48015 Gen 15 - Best score on D2: 0.21716 Best score on D1-D2 diff: 16.82297 Gen 16 - Best score on D2: 0.21716 Best score on D1-D2 diff: 16.82297 Gen 17 - Best score on D2: 0.21716 Best score on D1-D2 diff: 16.82297 Gen 18 - Best score on D2: 0.21716 Best score on D1-D2 diff: 20.37531 Gen 19 - Best score on D2: 0.21716 Best score on D1-D2 diff: 20.37531 Gen 20 - Best score on D2: 0.21716 Best score on D1-D2 diff: 20.37531 Gen 21 - Best score on D2: 0.21716 Best score on D1-D2 diff: 20.37531 Gen 22 - Best score on D2: 0.21716 Best score on D1-D2 diff: 20.37531 Gen 23 - Best score on D2: 0.21716 Best score on D1-D2 diff: 20.37531 Gen 24 - Best score on D2: 0.21716 Best score on D1-D2 diff: 20.37531 Gen 25 - Best score on D2: 0.21830 Best score on D1-D2 diff: 20.37531 ------------------------------------------------- Multiple Linear Regression: D2 Dataset R^2 trained on D1: -0.004100908941205184 Entire dataset(80%) R^2 trained on data (80%): 0.005604118099314936 Holdout R^2 (20%) trained on data (80%): -0.002302280932001022 Dataset D1 R^2 on trained D1: 0.007938814036301811 Combined Dataset (100%) R^2 trained on combined data (100%): 0.004902588418482234 ------------------------------------------------- Final Pareto Front: Pipeline #1: Score on D2: 0.21830056028037914 | D1-D2 diff: 1.5645984047111148 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), RecessiveEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=2, min_samples_split=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.38990818679169925 Holdout data R^2 trained on entire dataset(80%): 0.24420072515738256 Dataset D1 R^2 on trained D1: 0.3851744801374488 .................................................. Pipeline #2: Score on D2: 0.21716498969763187 | D1-D2 diff: 1.5972430191978266 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3736283736655093 Holdout data R^2 trained on entire dataset(80%): 0.24026322441621928 Dataset D1 R^2 on trained D1: 0.3708091312109253 .................................................. Pipeline #3: Score on D2: 0.21429711032622079 | D1-D2 diff: 1.6326318466013332 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=6, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3551936936448974 Holdout data R^2 trained on entire dataset(80%): 0.23486227305762652 Dataset D1 R^2 on trained D1: 0.3550466376707433 .................................................. Pipeline #4: Score on D2: 0.21237323171778033 | D1-D2 diff: 1.6847802397922675 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), RandomForestRegressor(max_features=0.25, min_samples_leaf=2, min_samples_split=20, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.343696987370664 Holdout data R^2 trained on entire dataset(80%): 0.2341000615596196 Dataset D1 R^2 on trained D1: 0.33648899981429226 .................................................. Pipeline #5: Score on D2: 0.2077413999689388 | D1-D2 diff: 1.7214679124638648 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.33034566655741815 Holdout data R^2 trained on entire dataset(80%): 0.2317624811450537 Dataset D1 R^2 on trained D1: 0.32161007621795545 .................................................. Pipeline #6: Score on D2: 0.2072257219838859 | D1-D2 diff: 1.7336500450426586 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=14, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.32453473482736295 Holdout data R^2 trained on entire dataset(80%): 0.22839270139217405 Dataset D1 R^2 on trained D1: 0.317927414066123 .................................................. Pipeline #7: Score on D2: 0.20691082041717357 | D1-D2 diff: 1.734120181822037 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.35000000000000003, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3209493326101468 Holdout data R^2 trained on entire dataset(80%): 0.22865941475436524 Dataset D1 R^2 on trained D1: 0.31749251210606344 .................................................. Pipeline #8: Score on D2: 0.2054748747683176 | D1-D2 diff: 1.8125796703417902 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3049656844165729 Holdout data R^2 trained on entire dataset(80%): 0.2254701114321802 Dataset D1 R^2 on trained D1: 0.2981176552914655 .................................................. Pipeline #9: Score on D2: 0.20409303087978548 | D1-D2 diff: 1.82685854893729 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.3054093294765672 Holdout data R^2 trained on entire dataset(80%): 0.22598323714838076 Dataset D1 R^2 on trained D1: 0.2938731779731074 .................................................. Pipeline #10: Score on D2: 0.20169816712483346 | D1-D2 diff: 1.8664395816226402 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2961062161560417 Holdout data R^2 trained on entire dataset(80%): 0.2272514655328095 Dataset D1 R^2 on trained D1: 0.2841014010738979 .................................................. Pipeline #11: Score on D2: 0.20001395999164262 | D1-D2 diff: 1.8843954944966843 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=15, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2916860593994244 Holdout data R^2 trained on entire dataset(80%): 0.2268740622054679 Dataset D1 R^2 on trained D1: 0.27932100553640604 .................................................. Pipeline #12: Score on D2: 0.199760510293434 | D1-D2 diff: 1.8852793223546795 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=13, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2865561933664543 Holdout data R^2 trained on entire dataset(80%): 0.21937516836185456 Dataset D1 R^2 on trained D1: 0.2789189423137821 .................................................. Pipeline #13: Score on D2: 0.19898351530272151 | D1-D2 diff: 1.890598941032624 Pipeline steps: HeterosisEncoder(), RecessiveEncoder(), HeterosisEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.3, min_samples_leaf=20, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29068681397029017 Holdout data R^2 trained on entire dataset(80%): 0.22428962928141782 Dataset D1 R^2 on trained D1: 0.2772547813786572 .................................................. Pipeline #14: Score on D2: 0.1989370896583259 | D1-D2 diff: 1.897633443543148 Pipeline steps: HeterosisEncoder(), RandomForestRegressor(max_features=0.45, min_samples_leaf=16, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.29273657052997204 Holdout data R^2 trained on entire dataset(80%): 0.22774610240997584 Dataset D1 R^2 on trained D1: 0.27605419104257645 .................................................. Pipeline #15: Score on D2: 0.19824019121651404 | D1-D2 diff: 1.9030305342072198 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.4, min_samples_leaf=16, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2851132832474803 Holdout data R^2 trained on entire dataset(80%): 0.22232821518990686 Dataset D1 R^2 on trained D1: 0.27448617520051777 .................................................. Pipeline #16: Score on D2: 0.19815847892055927 | D1-D2 diff: 1.9102677727720867 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2852587725900777 Holdout data R^2 trained on entire dataset(80%): 0.22527609459740083 Dataset D1 R^2 on trained D1: 0.2732555508743817 .................................................. Pipeline #17: Score on D2: 0.19760593907498802 | D1-D2 diff: 1.92319006066859 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(bootstrap=False, max_features=0.25, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2831711682556788 Holdout data R^2 trained on entire dataset(80%): 0.2198690399536084 Dataset D1 R^2 on trained D1: 0.27070489545794196 .................................................. Pipeline #18: Score on D2: 0.19674019785964025 | D1-D2 diff: 1.948672288290372 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=16, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.279929795424681 Holdout data R^2 trained on entire dataset(80%): 0.22147702288356352 Dataset D1 R^2 on trained D1: 0.2660899260586702 .................................................. Pipeline #19: Score on D2: 0.1961109758201881 | D1-D2 diff: 1.9675233488521042 Pipeline steps: HeterosisEncoder(), VarianceThreshold(), HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2762596171192736 Holdout data R^2 trained on entire dataset(80%): 0.21821621203424557 Dataset D1 R^2 on trained D1: 0.2628408675864776 .................................................. Pipeline #20: Score on D2: 0.19590425064419814 | D1-D2 diff: 2.0201058765570763 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=18, min_samples_split=12, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26995795063207595 Holdout data R^2 trained on entire dataset(80%): 0.22226678074361683 Dataset D1 R^2 on trained D1: 0.25595293158533206 .................................................. Pipeline #21: Score on D2: 0.19417579780501149 | D1-D2 diff: 2.024002759066753 Pipeline steps: SelectPercentile(percentile=95), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.35000000000000003, min_samples_leaf=16, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26742525380153004 Holdout data R^2 trained on entire dataset(80%): 0.21740495202963583 Dataset D1 R^2 on trained D1: 0.2537633573956959 .................................................. Pipeline #22: Score on D2: 0.19232555461080192 | D1-D2 diff: 2.048827844296746 Pipeline steps: HeterosisEncoder(), HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.3, min_samples_leaf=17, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.26632123530871 Holdout data R^2 trained on entire dataset(80%): 0.2168971967196618 Dataset D1 R^2 on trained D1: 0.24907715709568712 .................................................. Pipeline #23: Score on D2: 0.19153022896101835 | D1-D2 diff: 2.149888961997626 Pipeline steps: HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.05), RandomForestRegressor(max_features=0.3, min_samples_leaf=20, min_samples_split=15, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2552152361704242 Holdout data R^2 trained on entire dataset(80%): 0.21475075959980638 Dataset D1 R^2 on trained D1: 0.23833993138033627 .................................................. Pipeline #24: Score on D2: 0.18567778234880328 | D1-D2 diff: 2.1566005987128287 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24653139129538093 Holdout data R^2 trained on entire dataset(80%): 0.20980499909839934 Dataset D1 R^2 on trained D1: 0.2319074865074936 .................................................. Pipeline #25: Score on D2: 0.18447486686128822 | D1-D2 diff: 2.176793257414563 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24698810640605762 Holdout data R^2 trained on entire dataset(80%): 0.20644515955600662 Dataset D1 R^2 on trained D1: 0.2290129237627917 .................................................. Pipeline #26: Score on D2: 0.1832069020457585 | D1-D2 diff: 2.1802636785753653 Pipeline steps: VarianceThreshold(threshold=0.05), HeterosisEncoder(), OverDominanceEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24532176432193842 Holdout data R^2 trained on entire dataset(80%): 0.20751239287703804 Dataset D1 R^2 on trained D1: 0.2274620625905629 .................................................. Pipeline #27: Score on D2: 0.18122722043091388 | D1-D2 diff: 2.199022175338948 Pipeline steps: VarianceThreshold(threshold=0.05), HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=20, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.24574453604556434 Holdout data R^2 trained on entire dataset(80%): 0.20923357900949546 Dataset D1 R^2 on trained D1: 0.22399153981977127 .................................................. Pipeline #28: Score on D2: 0.17722215924825524 | D1-D2 diff: 2.2247689218190843 Pipeline steps: HeterosisEncoder(), UnderDominanceEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=17, min_samples_split=6, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23889020252163207 Holdout data R^2 trained on entire dataset(80%): 0.20169924516046356 Dataset D1 R^2 on trained D1: 0.21804097149129187 .................................................. Pipeline #29: Score on D2: 0.17496889485415268 | D1-D2 diff: 2.2409977316351166 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), HeterosisEncoder(), RandomForestRegressor(max_features=0.2, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2331702249117863 Holdout data R^2 trained on entire dataset(80%): 0.2020069536743725 Dataset D1 R^2 on trained D1: 0.21461808602112642 .................................................. Pipeline #30: Score on D2: 0.17059310033976915 | D1-D2 diff: 2.268104803283821 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), DominantEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2275750177725544 Holdout data R^2 trained on entire dataset(80%): 0.18936489706549042 Dataset D1 R^2 on trained D1: 0.2083805452245483 .................................................. Pipeline #31: Score on D2: 0.16857497394217658 | D1-D2 diff: 2.2969951956839947 Pipeline steps: OverDominanceEncoder(), SelectPercentile(percentile=70), DominantEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=15, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.23201417541577796 Holdout data R^2 trained on entire dataset(80%): 0.19441456610964258 Dataset D1 R^2 on trained D1: 0.20449690314847802 .................................................. Pipeline #32: Score on D2: 0.16846796380835327 | D1-D2 diff: 2.3148216148355423 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=15, min_samples_split=4, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.2271494107038211 Holdout data R^2 trained on entire dataset(80%): 0.19442173808749796 Dataset D1 R^2 on trained D1: 0.20329607193710075 .................................................. Pipeline #33: Score on D2: 0.16270318782028415 | D1-D2 diff: 2.4217887080114906 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), SelectPercentile(percentile=70), DominantEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=18, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20404241329224537 Holdout data R^2 trained on entire dataset(80%): 0.1672442851416488 Dataset D1 R^2 on trained D1: 0.1917738550912399 .................................................. Pipeline #34: Score on D2: 0.16214404293372597 | D1-D2 diff: 2.4273011868479557 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), SelectPercentile(percentile=70), DominantEncoder(), RandomForestRegressor(max_features=0.25, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20196832232525652 Holdout data R^2 trained on entire dataset(80%): 0.16736860001439524 Dataset D1 R^2 on trained D1: 0.1909515267797861 .................................................. Pipeline #35: Score on D2: 0.16011718316921597 | D1-D2 diff: 2.5030515304690333 Pipeline steps: HeterosisEncoder(), OverDominanceEncoder(), RandomForestRegressor(max_features=0.15000000000000002, min_samples_leaf=19, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.20990454014641924 Holdout data R^2 trained on entire dataset(80%): 0.18447138511209193 Dataset D1 R^2 on trained D1: 0.18559257296496778 .................................................. Pipeline #36: Score on D2: 0.14656304421267974 | D1-D2 diff: 2.5769605969699874 Pipeline steps: VarianceThreshold(threshold=0.35), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.2), VarianceThreshold(threshold=0.1), RandomForestRegressor(max_features=0.1, min_samples_leaf=16, min_samples_split=3, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.19936459291300757 Holdout data R^2 trained on entire dataset(80%): 0.16999510603455203 Dataset D1 R^2 on trained D1: 0.16923917159304824 .................................................. Pipeline #37: Score on D2: 0.1376027772545233 | D1-D2 diff: 2.815149716269294 Pipeline steps: SelectPercentile(percentile=95), HeterosisEncoder(), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=16, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.15200912793295263 Holdout data R^2 trained on entire dataset(80%): 0.14792011962314944 Dataset D1 R^2 on trained D1: 0.15352464505317587 .................................................. Pipeline #38: Score on D2: 0.136239799178697 | D1-D2 diff: 3.3129327262551933 Pipeline steps: UnderDominanceEncoder(), VarianceThreshold(threshold=0.25), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.14252665888541638 Holdout data R^2 trained on entire dataset(80%): 0.16304829303038515 Dataset D1 R^2 on trained D1: 0.14454116460525046 .................................................. Pipeline #39: Score on D2: 0.12954291480830116 | D1-D2 diff: 3.8060550160811397 Pipeline steps: SelectPercentile(percentile=95), HeterosisEncoder(), FeatureEncodingFrequencySelector(threshold=0.35), HeterosisEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=13, min_samples_split=8, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10298550722594224 Holdout data R^2 trained on entire dataset(80%): 0.09712306108918811 Dataset D1 R^2 on trained D1: 0.13430831911868724 .................................................. Pipeline #40: Score on D2: 0.1154257435164815 | D1-D2 diff: 20.375314103799774 Pipeline steps: UnderDominanceEncoder(), SelectPercentile(percentile=55), VarianceThreshold(threshold=0.25), UnderDominanceEncoder(), DecisionTreeRegressor(max_depth=4, min_samples_leaf=4, min_samples_split=9, random_state=42) Entire dataset(80%) R^2 trained on entire dataset(80%): 0.10354345650977981 Holdout data R^2 trained on entire dataset(80%): 0.09713746160398906 Dataset D1 R^2 on trained D1: 0.1154199414493422 ..................................................