/*********************************************************************** This file is part of KEEL-software, the Data Mining tool for regression, classification, clustering, pattern mining and so on. Copyright (C) 2004-2010 F. Herrera (herrera@decsai.ugr.es) L. S�nchez (luciano@uniovi.es) J. Alcal�-Fdez (jalcala@decsai.ugr.es) S. Garc�a (sglopez@ujaen.es) A. Fern�ndez (alberto.fernandez@ujaen.es) J. Luengo (julianlm@decsai.ugr.es) This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/ **********************************************************************/ package keel.Algorithms.Fuzzy_Rule_Learning.Genetic.Shared.Individual; import keel.Algorithms.Fuzzy_Rule_Learning.Shared.Fuzzy.*; import keel.Algorithms.Shared.Exceptions.*; /** * <p> * Class for management of genetic individuals in symbolic regression * </p> * * <p> * @author Written by Luciano S�nchez (University of Oviedo) 25/01/2004 * @author Modified by M.R. Su�rez (University of Oviedo) 18/12/2008 * @author Modified by Enrique A. de la Cal (University of Oviedo) 21/12/2008 * @version 1.0 * @since JDK1.5 * </p> */ public abstract class GeneticIndividualForSymbRegr extends GeneticIndividual { protected static FuzzyAlphaCut[][] Xfuzzy; protected static FuzzyAlphaCut[] Yfuzzy; protected static double[][] X; protected static double[] Y; protected static double[] Yo; protected FuzzyRegressor m; private static double ECM=0; private final static double MAXFIT=1e8; /** * <p> * Constructor. Initializes the type of fitness * </p> * @param tf the type of fitness */ public GeneticIndividualForSymbRegr(int tf) { super(tf); } /** * <p> * This method calculates the fitness based in the ECM * </p> * @return The fitness * @throws invalidFitness Message is error */ public double fitness() throws invalidFitness { // if (tipoFitness == STANDARD) { FuzzyAlphaCut fECM = new FuzzyAlphaCut(new FuzzyNumberTRIANG(0,0,0)); for (int i=0;i<X.length;i++) { FuzzyAlphaCut obtainedOutput=m.output(Xfuzzy[i]); FuzzyAlphaCut fERROR = obtainedOutput.subtract(Yfuzzy[i]); fECM = fECM.sum(fERROR.sqr()); } fECM = fECM.multiply(1.0/X.length); return fECM.massCentre(); } /** * <p> * This method is for debug the fitness * </p> */ public void debug_fitness() { double ECMT=0; for (int i=0;i<X.length;i++) { FuzzyAlphaCut obtainedOutput=m.output(Xfuzzy[i]); double y=obtainedOutput.massCentre(); double error=y-Y[i]; ECMT+=error*error; } ECMT/=X.length; System.out.println(" Error cuadratico medio defuzzificado="+ECMT); } /** * <p> * This method is for debug * </p> */ public void debug() { g.debug(); m.debug(); } /** * <p> * This method assign examples based on a level of tolerance * </p> * @param pX * @param pY * @param tolerance The level of tolerance */ public void asignaejemplos(double[][] pX, double[] pY, double tolerance) { X=pX; Y=pY; Xfuzzy=new FuzzyAlphaCut[pX.length][]; Yfuzzy=new FuzzyAlphaCut[pY.length]; for (int i=0;i<pX.length;i++) { Xfuzzy[i] = new FuzzyAlphaCut[pX[i].length]; for (int j=0;j<X[i].length;j++) Xfuzzy[i][j]=new FuzzyAlphaCut(new FuzzyNumberTRIANG(pX[i][j]*(1-tolerance),pX[i][j],pX[i][j]*(1+tolerance))); Yfuzzy[i]=new FuzzyAlphaCut(new FuzzyNumberTRIANG(pY[i],pY[i],pY[i])); } } /** * <p> * This method obtain a crips output that we can compare to punctual models * </p> * @return The crisp output */ public double[] getYo() { // Obtains a crisp output that we can compare to // punctual models Yo = new double[X.length]; for (int i=0;i<X.length;i++) { FuzzyAlphaCut obtainedOutput=m.output(Xfuzzy[i]); Yo[i]=obtainedOutput.massCentre(); } return Yo; } /** * <p> * This method calculate the mean square error * @return The mean square error */ public double MSE() { // Mean Square Error of mass center output is calculated on the examples double error=0; double sumY=0; for (int i=0;i<X.length;i++) { FuzzyAlphaCut obtainedOutput=m.output(Xfuzzy[i]); error=Math.abs(Y[i]-obtainedOutput.massCentre()); sumY+=Y[i]; ECM+=error*error; } ECM/=X.length; return ECM; } }