Core.BaseLearners.LearningAttributeWeights

Class LearningAttributeWeights

java.lang.Object

Core.BaseLearners.LearningAttributeWeights.LearningAttributeWeights

Direct Known Subclasses:

GDweightLearning, GeneticWeightLearning, GWeightingProcessLearning, LearnNot, LRSweightLearning, PSOweightLearning, PSOweightLearningPEPPER, SalzbergModified, SalzbergWeightLearning, SAweightLearning

public abstract class LearningAttributeWeights
extends java.lang.Object

Classe generica que permet fer 'Learning' dels pesos dels atributs de la base de casos del CBR. Es un classe abstracta que declara els metodes que s'hauran fer servir en les subclasses.

Field Summary

Fields

Modifier and Type	Field and Description
protected java.lang.String	path

Constructor Summary

Constructors

Constructor and Description
LearningAttributeWeights()

Method Summary

Methods

Modifier and Type	Method and Description
protected float[]	addIgnoredWeights(float[] weights) dropIgnored must have been called before this method.
abstract Attribute[]	learn(CBR cbr, Attribute[] attributes) Mètode de aprenentatge de pesos que modifica el pes dels atributs de la Population train en funció de la Population de test
Attribute[]	learnWeights(CBR cbr, Attribute[] attributes)

Methods inherited from class java.lang.Object

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

Field Detail

path

protected java.lang.String path

Constructor Detail

LearningAttributeWeights

public LearningAttributeWeights()

Method Detail

learn

public abstract Attribute[] learn(CBR cbr,
                Attribute[] attributes)

Mètode de aprenentatge de pesos que modifica el pes dels atributs de la Population train en funció de la Population de test

Parameters:

cbr -

attributes -

Returns:

learnWeights

public Attribute[] learnWeights(CBR cbr,
                       Attribute[] attributes)

addIgnoredWeights

protected float[] addIgnoredWeights(float[] weights)

dropIgnored must have been called before this method. This method adds the weights of the dropped attributes to a given array of non-ignored attribute weights into their original position. This method must be called whenever weights are to be set in a cbr, or else the weights lengths won't match.

Parameters:

weights - float array with the weights of non-ignored attributes

Returns:

float array with the weights of dropped and non-dropped attributes