meka.classifiers.multilabel

Interface SemisupervisedClassifier

All Superinterfaces:

weka.classifiers.Classifier, MultiLabelClassifier, weka.core.OptionHandler

All Known Implementing Classes:

CM, EM

public interface SemisupervisedClassifier
extends MultiLabelClassifier

SemisupervisedClassifier.java - An Interface for Multilabel Semisupervised Classifiers. This is an interface for multi-label semisupervised classificatation, i.e., training on a partially labelled dataset.
For classifiers implementing this interface, the method introduceUnlabelledData(unlabeledInstances) will be called prior to buildClassifier(trainingInstances).
As of writing this comment, the unlabelled data comes only from the test data -- there is not yet any option for setting a seperate sete of unlabelled data (although this is planned for future versions).

Version:

September 2015

Author:

Jesse Read

Method Summary

Modifier and Type	Method and Description
void	introduceUnlabelledData(weka.core.Instances unlabeledInstances) Set Unlabelled Data - provide unlabelled data prior to calling buildClassifier(Instances).

Methods inherited from interface meka.classifiers.multilabel.MultiLabelClassifier

debugTipText, getDebug, getModel, setDebug

Methods inherited from interface weka.classifiers.Classifier

buildClassifier, classifyInstance, distributionForInstance, getCapabilities

Methods inherited from interface weka.core.OptionHandler

getOptions, listOptions, setOptions

Method Detail

introduceUnlabelledData

void introduceUnlabelledData(weka.core.Instances unlabeledInstances)

Set Unlabelled Data - provide unlabelled data prior to calling buildClassifier(Instances).

Parameters:

unlabeledInstances - Instances for which the true class labels are not available for each instance.