net.sf.javaml.featureselection.scoring
Class SymmetricalUncertainty

java.lang.Object
  net.sf.javaml.featureselection.scoring.SymmetricalUncertainty

All Implemented Interfaces:

FeatureScoring, FeatureSelection

public class SymmetricalUncertainty
extends java.lang.Object
implements FeatureScoring

Implements the Symmetrical Uncertainty (SU) evaluation method for attributes.

Version:

%SVN.VERSION%

Author:

Thomas Abeel (thomas@abeel.be), Mark Hall

Constructor Summary

SymmetricalUncertainty()

Method Summary

void	build(Dataset data) Build the attribute evaluation on the supplied data set.
int	noAttributes() Returns the number of attributes that have been ranked, scored or selected.
double	score(int attribute) Evaluates an individual attribute by measuring the symmetrical uncertainty between it and the class.

Methods inherited from class java.lang.Object

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

Constructor Detail

SymmetricalUncertainty

public SymmetricalUncertainty()

Method Detail

build

public void build(Dataset data)

Description copied from interface: FeatureSelection

Build the attribute evaluation on the supplied data set. Note: This method can change the data set that is supplied to the method!

Specified by:

build in interface FeatureSelection

Parameters:

data - data set to train the attribute evaluation algorithm on.

score

public double score(int attribute)

Evaluates an individual attribute by measuring the symmetrical uncertainty between it and the class.

Specified by:

score in interface FeatureScoring

Parameters:

attribute - the index of the attribute to be evaluated

Returns:

the symmetrical uncertainty

noAttributes

public int noAttributes()

Description copied from interface: FeatureSelection

Returns the number of attributes that have been ranked, scored or selected.

Specified by:

noAttributes in interface FeatureSelection

Returns:

the number of ranked, scored or selected attributes