uk.ac.ed.inf.csb.taweka.mulan.learn
Class MulanEvaluator

java.lang.Object
  mulan.evaluation.Evaluator
      uk.ac.ed.inf.csb.taweka.mulan.learn.MulanEvaluator

public class MulanEvaluator
extends mulan.evaluation.Evaluator

Class created to extract individual evaluation results (not possible from the Mulan class Evaluator, because only averaged results are returned).

Version:

8 Jun 2010

Author:

Luna De Ferrari luna.deferrari-at-ed.ac.uk

Field Summary

Fields inherited from class mulan.evaluation.Evaluator

DEFAULTFOLDS

Constructor Summary

MulanEvaluator(java.lang.String arffFilePath, java.lang.String xmlFilePath, mulan.classifier.MultiLabelLearner learner)

Method Summary

mulan.evaluation.Evaluation[]	crossValidate(mulan.classifier.MultiLabelLearner learner, mulan.data.MultiLabelInstances mlDataSet, int numFolds, int seed) Evaluates a MultiLabelLearner via cross-validation on given data set with defined number of folds.
java.lang.String	evaluate(mulan.classifier.MultiLabelLearner learner, mulan.data.MultiLabelInstances mlDataSet, int numFolds) Evaluates (runs a cross validation)
mulan.classifier.MultiLabelLearner	getAlgorithm()
java.lang.String	getArffFilePath()
java.lang.String	getEndTime()
mulan.evaluation.Evaluation[]	getEvaluationResults()
java.lang.String	getStartTime()
java.lang.String	getXmlFilePath()

Methods inherited from class mulan.evaluation.Evaluator

crossValidate, crossValidate, crossValidate, crossValidate, evaluate, evaluate

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

MulanEvaluator

public MulanEvaluator(java.lang.String arffFilePath,
                      java.lang.String xmlFilePath,
                      mulan.classifier.MultiLabelLearner learner)

Parameters:

arffFilePath -

xmlFilePath -

learner -

Method Detail

crossValidate

public mulan.evaluation.Evaluation[] crossValidate(mulan.classifier.MultiLabelLearner learner,
                                                   mulan.data.MultiLabelInstances mlDataSet,
                                                   int numFolds,
                                                   int seed)
                                            throws java.lang.Exception

Evaluates a MultiLabelLearner via cross-validation on given data set with defined number of folds. The specified number of folds has to be at least two. If negative value is specified, the used number of folds is equal to number of instances in the data set.

Parameters:

learner - the learner to be evaluated via cross-validation

mlDataSet - the multi-label data set for cross-validation

numFolds - the number of folds to be used

Returns:

the evaluation results (one Evaluation per fold combination)

Throws:

java.lang.IllegalArgumentException - if either of learner or data set parameters is null

java.lang.IllegalArgumentException - if number of folds is invalid

java.lang.Exception

evaluate

public java.lang.String evaluate(mulan.classifier.MultiLabelLearner learner,
                                 mulan.data.MultiLabelInstances mlDataSet,
                                 int numFolds)

Evaluates (runs a cross validation)

Parameters:

learner -

mlDataSet -

numFolds -

Returns:

the timestamp of the end of the computation

getAlgorithm

public mulan.classifier.MultiLabelLearner getAlgorithm()

getArffFilePath

public java.lang.String getArffFilePath()

getEndTime

public java.lang.String getEndTime()

getEvaluationResults

public mulan.evaluation.Evaluation[] getEvaluationResults()

getStartTime

public java.lang.String getStartTime()

getXmlFilePath

public java.lang.String getXmlFilePath()