moa.classifiers.core.splitcriteria

Interface SplitCriterion

All Superinterfaces:

MOAObject, OptionHandler, Serializable

All Known Implementing Classes:

GiniSplitCriterion, InfoGainSplitCriterion, InfoGainSplitCriterionMultilabel, SDRSplitCriterion, SDRSplitCriterionAMRules, VarianceReductionSplitCriterion

public interface SplitCriterion
extends OptionHandler

Interface for computing splitting criteria. with respect to distributions of class values. The split criterion is used as a parameter on decision trees and decision stumps. The two split criteria most used are Information Gain and Gini.

Version:

$Revision: 7 $

Author:

Richard Kirkby (rkirkby@cs.waikato.ac.nz)

Method Summary

Methods

Modifier and Type	Method and Description
double	getMeritOfSplit(double[] preSplitDist, double[][] postSplitDists) Computes the merit of splitting for a given ditribution before the split and after it.
double	getRangeOfMerit(double[] preSplitDist) Computes the range of splitting merit

Methods inherited from interface moa.options.OptionHandler

copy, getCLICreationString, getOptions, getPurposeString, prepareForUse, prepareForUse

Methods inherited from interface moa.MOAObject

getDescription, measureByteSize

Method Detail

getMeritOfSplit

double getMeritOfSplit(double[] preSplitDist,
                     double[][] postSplitDists)

Computes the merit of splitting for a given ditribution before the split and after it.

Parameters:

preSplitDist - the class distribution before the split

postSplitDist - the class distribution after the split

Returns:

value of the merit of splitting

getRangeOfMerit

double getRangeOfMerit(double[] preSplitDist)

Computes the range of splitting merit

Parameters:

preSplitDist - the class distribution before the split

Returns:

value of the range of splitting merit