Package weka.classifiers

Interface AbstainingClassifier

All Superinterfaces:

weka.classifiers.Classifier

All Known Implementing Classes:

AbstainAttributePercentile, AbstainAverage, AbstainAverageWithClassifierWeights, AbstainingClassifierWrapper, AbstainingLWL, AbstainLeastMedianSq, AbstainMinimumProbability, AbstainVote, Consensus, ConsensusOrVote, FilteredClassifierExt, MinMaxLimits, ThreadSafeClassifierWrapper

public interface AbstainingClassifier
extends weka.classifiers.Classifier

Interface for classifiers that may support abstaining.

Version:

$Revision$

Author:

fracpete (fracpete at waikato dot ac dot nz)

Method Summary

Modifier and Type	Method	Description
boolean	canAbstain()	Whether abstaining is possible, e.g., used in meta-classifiers.
double	getAbstentionClassification(weka.core.Instance inst)	The prediction that made the classifier abstain.
double[]	getAbstentionDistribution(weka.core.Instance inst)	The class distribution that made the classifier abstain.

Methods inherited from interface weka.classifiers.Classifier

buildClassifier, classifyInstance, distributionForInstance, getCapabilities

Method Detail

canAbstain

boolean canAbstain()

Whether abstaining is possible, e.g., used in meta-classifiers.

Returns:

true if abstaining is possible

getAbstentionClassification

double getAbstentionClassification(weka.core.Instance inst)
                            throws Exception

The prediction that made the classifier abstain.

Parameters:

inst - the instance to get the prediction for

Returns:

the prediction

Throws:

Exception - if fails to make prediction

getAbstentionDistribution

double[] getAbstentionDistribution(weka.core.Instance inst)
                            throws Exception

The class distribution that made the classifier abstain.

Parameters:

inst - the instance to get the prediction for

Returns:

the class distribution

Throws:

Exception - if fails to make prediction