keel.Algorithms.Genetic_Rule_Learning.MPLCS

Class ClassifierADI

java.lang.Object

keel.Algorithms.Genetic_Rule_Learning.MPLCS.Classifier

keel.Algorithms.Genetic_Rule_Learning.MPLCS.ClassifierADI

All Implemented Interfaces:

java.lang.Cloneable

public class ClassifierADI
extends Classifier
implements java.lang.Cloneable

Contains the classifier for the ADI knowledge representation

Since:

JDK1.2

Version:

1.1

Author:

Written by Jaume Bacardit (La Salle, Ramón Llull University - Barcelona) 28/03/2004, Modified by Xavi Solé (La Salle, Ramón Llull University - Barcelona) 23/12/2008, Modified by Jose A. Saez Munoz (ETSIIT, Universidad de Granada - Granada) 10/09/10

Field Summary

Fields inherited from class keel.Algorithms.Genetic_Rule_Learning.MPLCS.Classifier

accuracy, bloatControlDone, exceptionsLength, fitness, isEvaluated, numAliveRules, numRules, positionRuleMatch, theoryLength

Constructor Summary

Constructors

Constructor and Description
ClassifierADI()

Method Summary

Modifier and Type	Method and Description
double	computeTheoryLength()
Classifier	copy()
Classifier[]	crossoverClassifiers(Classifier _parent2)
void	crossoverRSW(Classifier[] parents, int num)
void	deleteRules(int[] whichRules)
void	doLocalSearch()
int	doMatch(InstanceWrapper ins) This function classifies input instances.
void	doMerge()
void	doMutation()
void	doReinitialize()
void	doSpecialStage(int stage)
void	doSplit()
matchProfileAgent	evaluateRuleCX(int[] rule)
void	findOrder(java.util.Vector<matchProfileAgent> profiles, java.util.Vector<int[]> candidateRules, java.util.Vector<java.lang.Integer> ruleOrder)
double	getLength()
int	getNiche()
int	getNumNiches()
int	getNumRules()
void	initRandomClassifier()
int	numSpecialStages()
void	printClassifier()

Methods inherited from class keel.Algorithms.Genetic_Rule_Learning.MPLCS.Classifier

compareToIndividual, computePerformance, getExceptionsLength, getFitness, getIsEvaluated, getNumAliveRules, getPositionRuleMatch, getTheoryLength, resetPerformance, setAccuracy, setExceptionsLength, setFitness, setIsEvaluated, setNumAliveRules, setPositionRuleMatch

Methods inherited from class java.lang.Object

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

Constructor Detail

ClassifierADI

public ClassifierADI()

Method Detail

initRandomClassifier

public void initRandomClassifier()

Specified by:

initRandomClassifier in class Classifier

computeTheoryLength

public double computeTheoryLength()

Specified by:

computeTheoryLength in class Classifier

getLength

public double getLength()

Specified by:

getLength in class Classifier

doMatch

public int doMatch(InstanceWrapper ins)

This function classifies input instances. It returns a class prediction of -1 if the input example cannot be classified

Specified by:

doMatch in class Classifier

printClassifier

public void printClassifier()

Specified by:

printClassifier in class Classifier

getNumRules

public int getNumRules()

Specified by:

getNumRules in class Classifier

crossoverClassifiers

public Classifier[] crossoverClassifiers(Classifier _parent2)

Specified by:

crossoverClassifiers in class Classifier

copy

public Classifier copy()

Specified by:

copy in class Classifier

doMutation

public void doMutation()

Specified by:

doMutation in class Classifier

deleteRules

public void deleteRules(int[] whichRules)

Specified by:

deleteRules in class Classifier

numSpecialStages

public int numSpecialStages()

Specified by:

numSpecialStages in class Classifier

doSpecialStage

public void doSpecialStage(int stage)

Specified by:

doSpecialStage in class Classifier

doSplit

public void doSplit()

doMerge

public void doMerge()

doReinitialize

public void doReinitialize()

getNiche

public int getNiche()

Specified by:

getNiche in class Classifier

getNumNiches

public int getNumNiches()

Specified by:

getNumNiches in class Classifier

doLocalSearch

public void doLocalSearch()

Specified by:

doLocalSearch in class Classifier

crossoverRSW

public void crossoverRSW(Classifier[] parents,
                         int num)

Specified by:

crossoverRSW in class Classifier

evaluateRuleCX

public matchProfileAgent evaluateRuleCX(int[] rule)

findOrder

public void findOrder(java.util.Vector<matchProfileAgent> profiles,
                      java.util.Vector<int[]> candidateRules,
                      java.util.Vector<java.lang.Integer> ruleOrder)