it.uniroma2.sag.kelp.learningalgorithm.classification.multiclassification

Class OneVsOneLearning

java.lang.Object

it.uniroma2.sag.kelp.learningalgorithm.classification.multiclassification.OneVsOneLearning

All Implemented Interfaces:

ClassificationLearningAlgorithm, LearningAlgorithm, MetaLearningAlgorithm

public class OneVsOneLearning
extends Object
implements ClassificationLearningAlgorithm, MetaLearningAlgorithm

It is a meta algorithm that operates by applying a One-Vs-One strategy over the base learning algorithm which is intended to be a binary learner. This meta-algorithms will learn N*(N-1)/2 classifiers, by comparing each class with all the others separately. The resulting classifier applies a a voting stratefy to perform the final decision. (N is the number of classes in the dataset)

NOTE: the base learning algorithm must provide a duplicate method which properly works

Author:

Giuseppe Castellucci, Simone Filice

Constructor Summary

Constructors

Constructor and Description
OneVsOneLearning()

Method Summary

Methods

Modifier and Type	Method and Description
OneVsOneLearning	duplicate() This method will duplicate the current Learning algorithm
LearningAlgorithm	getBaseAlgorithm() This method will return the base algorithm.
List<Label>	getLabels() Returns the labels to be learned applying a one-vs-one strategy
OneVsOneClassifier	getPredictionFunction() This method returns the learned PredictionFunction.
void	learn(Dataset dataset) This method will cause the meta-learning algorithm to learn N*(N-1)/2 classifiers, where N is the number of classes in the dataset.
void	reset() This method will cause the reset of all the base algorithms
void	setBaseAlgorithm(LearningAlgorithm baseAlgorithm) This method will set the type of the base algorithms to be learned.
void	setLabels(List<Label> labels) Set the labels associated to this multi-classifier.

Methods inherited from class java.lang.Object

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

Constructor Detail

OneVsOneLearning

public OneVsOneLearning()

Method Detail

setLabels

public void setLabels(List<Label> labels)

Set the labels associated to this multi-classifier.

Specified by:

setLabels in interface LearningAlgorithm

Parameters:

labels - the labels representing the concept to be learned

getLabels

public List<Label> getLabels()

Returns the labels to be learned applying a one-vs-one strategy

Specified by:

getLabels in interface LearningAlgorithm

Returns:

the labels to be learned

learn

public void learn(Dataset dataset)

This method will cause the meta-learning algorithm to learn N*(N-1)/2 classifiers, where N is the number of classes in the dataset.

Specified by:

learn in interface LearningAlgorithm

Parameters:

dataset - the training data

reset

public void reset()

This method will cause the reset of all the base algorithms

Specified by:

reset in interface LearningAlgorithm

getPredictionFunction

public OneVsOneClassifier getPredictionFunction()

This method returns the learned PredictionFunction. It is of type OneVsOneClassifier.

Specified by:

getPredictionFunction in interface ClassificationLearningAlgorithm

Specified by:

getPredictionFunction in interface LearningAlgorithm

Returns:

the predictionFunction learned during the training process

setBaseAlgorithm

public void setBaseAlgorithm(LearningAlgorithm baseAlgorithm)

This method will set the type of the base algorithms to be learned.

Specified by:

setBaseAlgorithm in interface MetaLearningAlgorithm

Parameters:

baseAlgorithm - the baseAlgorithm to set

getBaseAlgorithm

public LearningAlgorithm getBaseAlgorithm()

This method will return the base algorithm.

Specified by:

getBaseAlgorithm in interface MetaLearningAlgorithm

Returns:

the base algorithm

duplicate

public OneVsOneLearning duplicate()

This method will duplicate the current Learning algorithm

Specified by:

duplicate in interface LearningAlgorithm