com.rapidminer.operator.learner.tree
Interface Criterion

All Known Implementing Classes:

AbstractCriterion, AccuracyCriterion, GainRatioCriterion, GiniIndexCriterion, InfoGainCriterion

public interface Criterion

The criterion for a splitted example set. Possible implementations are for example accuracy or information gain.

Author:

Sebastian Land, Ingo Mierswa

Method Summary

double	getIncrementalBenefit()
double	getNominalBenefit(ExampleSet exampleSet, Attribute attribute)
double	getNumericalBenefit(ExampleSet exampleSet, Attribute attribute, double splitValue)
void	startIncrementalCalculation(ExampleSet exampleSet)
boolean	supportsIncrementalCalculation()
void	swapExample(Example example)

Method Detail

getNominalBenefit

double getNominalBenefit(ExampleSet exampleSet,
                         Attribute attribute)

getNumericalBenefit

double getNumericalBenefit(ExampleSet exampleSet,
                           Attribute attribute,
                           double splitValue)

supportsIncrementalCalculation

boolean supportsIncrementalCalculation()

startIncrementalCalculation

void startIncrementalCalculation(ExampleSet exampleSet)

swapExample

void swapExample(Example example)

getIncrementalBenefit

double getIncrementalBenefit()