/***********************************************************************
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/
**********************************************************************/
/**
* <p>
* File: SMOTE_ENN.java
* </p>
*
* The SMOTE ENN algorithm is an oversampling method used to deal with
* the imbalanced problem.
*
* @author Written by Salvador Garcia Lopez (University of Granada) 30/03/2006
* @author Modified by Victoria Lopez Morales (University of Granada) 21/09/2010
* @version 0.1
* @since JDK1.5
*
*/
package keel.Algorithms.ImbalancedClassification.Resampling.SMOTE_ENN;
import keel.Algorithms.Preprocess.Basic.*;
import keel.Dataset.Attribute;
import keel.Dataset.Attributes;
import keel.Dataset.Instance;
import org.core.*;
import java.util.StringTokenizer;
public class SMOTE_ENN extends Metodo {
/**
* <p>
* The SMOTE ENN algorithm is an oversampling method used to deal with
* the imbalanced problem.
* </p>
*/
/*Own parameters of the algorithm*/
private long semilla;
private int kSMOTE;
private int kENN;
private int ASMO;
private boolean balance;
private double smoting;
/**
* <p>
* Constructor of the class. It configures the execution of the algorithm by
* reading the configuration script that indicates the parameters that are
* going to be used.
* </p>
*
* @param ficheroScript Name of the configuration script that indicates the
* parameters that are going to be used during the execution of the algorithm
*/
public SMOTE_ENN (String ficheroScript) {
super (ficheroScript);
}
/**
* <p>
* The main method of the class that includes the operations of the algorithm.
* It includes all the operations that the algorithm has and finishes when it
* writes the output information into files.
* </p>
*/
public void run () {
int nPos = 0;
int nNeg = 0;
int i, j, l, m;
int tmp;
int posID, negID;
int positives[];
double conjS[][];
double conjR[][];
int conjN[][];
boolean conjM[][];
int clasesS[];
double genS[][];
double genR[][];
int genN[][];
boolean genM[][];
int clasesGen[];
int tamS;
int pos;
int neighbors[][];
int nn;
boolean marcas[];
int nSel = 0, claseObt;
double conjS2[][];
double conjR2[][];
int conjN2[][];
boolean conjM2[][];
int clasesS2[];
long tiempo = System.currentTimeMillis();
/*SMOTE PART*/
/*Count of number of positive and negative examples*/
for (i=0; i<clasesTrain.length; i++) {
if (clasesTrain[i] == 0)
nPos++;
else
nNeg++;
}
if (nPos > nNeg) {
tmp = nPos;
nPos = nNeg;
nNeg = tmp;
posID = 1;
negID = 0;
} else {
posID = 0;
negID = 1;
}
/*Localize the positive instances*/
positives = new int[nPos];
for (i=0, j=0; i<clasesTrain.length; i++) {
if (clasesTrain[i] == posID) {
positives[j] = i;
j++;
}
}
/*Randomize the instance presentation*/
Randomize.setSeed (semilla);
for (i=0; i<positives.length; i++) {
tmp = positives[i];
pos = Randomize.Randint(0,positives.length-1);
positives[i] = positives[pos];
positives[pos] = tmp;
}
/*Obtain k-nearest neighbors of each positive instance*/
neighbors = new int[positives.length][kSMOTE];
for (i=0; i<positives.length; i++) {
switch (ASMO) {
case 0:
KNN.evaluacionKNN2 (kSMOTE, datosTrain, realTrain, nominalTrain, nulosTrain, clasesTrain, datosTrain[positives[i]], realTrain[positives[i]], nominalTrain[positives[i]], nulosTrain[positives[i]], 2, distanceEu, neighbors[i]);
break;
case 1:
evaluationKNNClass (kSMOTE, datosTrain, realTrain, nominalTrain, nulosTrain, clasesTrain, datosTrain[positives[i]], realTrain[positives[i]], nominalTrain[positives[i]], nulosTrain[positives[i]], 2, distanceEu, neighbors[i],posID);
break;
case 2:
evaluationKNNClass (kSMOTE, datosTrain, realTrain, nominalTrain, nulosTrain, clasesTrain, datosTrain[positives[i]], realTrain[positives[i]], nominalTrain[positives[i]], nulosTrain[positives[i]], 2, distanceEu, neighbors[i],negID);
break;
}
}
/*Interpolation of the minority instances*/
if (balance) {
genS = new double[nNeg-nPos][datosTrain[0].length];
genR = new double[nNeg-nPos][datosTrain[0].length];
genN = new int[nNeg-nPos][datosTrain[0].length];
genM = new boolean[nNeg-nPos][datosTrain[0].length];
clasesGen = new int[nNeg-nPos];
} else {
genS = new double[(int)(nPos*smoting)][datosTrain[0].length];
genR = new double[(int)(nPos*smoting)][datosTrain[0].length];
genN = new int[(int)(nPos*smoting)][datosTrain[0].length];
genM = new boolean[(int)(nPos*smoting)][datosTrain[0].length];
clasesGen = new int[(int)(nPos*smoting)];
}
for (i=0; i<genS.length; i++) {
clasesGen[i] = posID;
nn = Randomize.Randint(0,kSMOTE-1);
interpolate (realTrain[positives[i%positives.length]],realTrain[neighbors[i%positives.length][nn]],nominalTrain[positives[i%positives.length]],nominalTrain[neighbors[i%positives.length][nn]],nulosTrain[positives[i%positives.length]],nulosTrain[neighbors[i%positives.length][nn]],genS[i],genR[i],genN[i],genM[i]);
}
if (balance) {
tamS = 2*nNeg;
} else {
tamS = nNeg + nPos + (int)(nPos*smoting);
}
/*Construction of the S set from the previous vector S*/
conjS = new double[tamS][datosTrain[0].length];
conjR = new double[tamS][datosTrain[0].length];
conjN = new int[tamS][datosTrain[0].length];
conjM = new boolean[tamS][datosTrain[0].length];
clasesS = new int[tamS];
for (j=0; j<datosTrain.length; j++) {
for (l=0; l<datosTrain[0].length; l++) {
conjS[j][l] = datosTrain[j][l];
conjR[j][l] = realTrain[j][l];
conjN[j][l] = nominalTrain[j][l];
conjM[j][l] = nulosTrain[j][l];
}
clasesS[j] = clasesTrain[j];
}
for (m=0;j<tamS; j++, m++) {
for (l=0; l<datosTrain[0].length; l++) {
conjS[j][l] = genS[m][l];
conjR[j][l] = genR[m][l];
conjN[j][l] = genN[m][l];
conjM[j][l] = genM[m][l];
}
clasesS[j] = clasesGen[m];
}
/*ENN PART*/
/*Inicialization of the flagged instances vector for a posterior copy*/
marcas = new boolean[conjS.length];
for (i=0; i<conjS.length; i++)
marcas[i] = false;
/*Body of the algorithm. For each instance in T, search the correspond class conform his mayority
from the nearest neighborhood. Is it is positive, the instance is selected.*/
for (i=0; i<conjS.length; i++) {
/*Apply KNN to the instance*/
claseObt = KNN.evaluacionKNN2 (kENN, conjS, clasesS, conjS[i], 2);
if (claseObt == clasesS[i]) { //conform with your mayority, it is included in the solution set
marcas[i] = true;
nSel++;
}
}
/*Building of the S set from the flags*/
conjS2 = new double[nSel][conjS[0].length];
conjR2 = new double[nSel][conjS[0].length];
conjN2 = new int[nSel][conjS[0].length];
conjM2 = new boolean[nSel][conjS[0].length];
clasesS2 = new int[nSel];
for (i=0, l=0; i<conjS.length; i++) {
if (marcas[i]) { //the instance will be copied to the solution
for (j=0; j<conjS[0].length; j++) {
conjS2[l][j] = conjS[i][j];
conjR2[l][j] = conjR[i][j];
conjN2[l][j] = conjN[i][j];
conjM2[l][j] = conjM[i][j];
}
clasesS2[l] = clasesS[i];
l++;
}
}
System.out.println("SMOTE_ENN "+ relation + " " + (double)(System.currentTimeMillis()-tiempo)/1000.0 + "s");
OutputIS.escribeSalida(ficheroSalida[0], conjR2, conjN2, conjM2, clasesS2, entradas, salida, nEntradas, relation);
OutputIS.escribeSalida(ficheroSalida[1], test, entradas, salida, nEntradas, relation);
}
/**
* <p>
* Computes the k nearest neighbors of a given item belonging to a fixed class.
* With that neighbors a suggested class for the item is returned.
* </p>
*
* @param nvec Number of nearest neighbors that are going to be searched
* @param conj Matrix with the data of all the items in the dataset
* @param real Matrix with the data associated to the real attributes of the dataset
* @param nominal Matrix with the data associated to the nominal attributes of the dataset
* @param nulos Matrix with the data associated to the missing values of the dataset
* @param clases Array with the associated class for each item in the dataset
* @param ejemplo Array with the data of the specific item in the dataset used
* as a reference in the nearest neighbor search
* @param ejReal Array with the data of the real attributes of the specific item in the dataset
* @param ejNominal Array with the data of the nominal attributes of the specific item in the dataset
* @param ejNulos Array with the data of the missing values of the specific item in the dataset
* @param nClases Class of the specific item in the dataset
* @param distance Kind of distance used in the nearest neighbors computation.
* If true the distance used is the euclidean, if false the HVMD distance is used
* @param vecinos Array that will have the nearest neighbours id for the current specific item
* @param clase Class of the neighbours searched for the item
* @return the majority class for all the neighbors of the item
*/
public static int evaluationKNNClass (int nvec, double conj[][], double real[][], int nominal[][], boolean nulos[][], int clases[], double ejemplo[], double ejReal[], int ejNominal[], boolean ejNulos[], int nClases, boolean distance, int vecinos[], int clase) {
int i, j, l;
boolean parar = false;
int vecinosCercanos[];
double minDistancias[];
int votos[];
double dist;
int votada, votaciones;
if (nvec > conj.length)
nvec = conj.length;
votos = new int[nClases];
vecinosCercanos = new int[nvec];
minDistancias = new double[nvec];
for (i=0; i<nvec; i++) {
vecinosCercanos[i] = -1;
minDistancias[i] = Double.POSITIVE_INFINITY;
}
for (i=0; i<conj.length; i++) {
dist = KNN.distancia(conj[i], real[i], nominal[i], nulos[i], ejemplo, ejReal, ejNominal, ejNulos, distance);
if (dist > 0 && clases[i] == clase) {
parar = false;
for (j = 0; j < nvec && !parar; j++) {
if (dist < minDistancias[j]) {
parar = true;
for (l = nvec - 1; l >= j+1; l--) {
minDistancias[l] = minDistancias[l - 1];
vecinosCercanos[l] = vecinosCercanos[l - 1];
}
minDistancias[j] = dist;
vecinosCercanos[j] = i;
}
}
}
}
for (j=0; j<nClases; j++) {
votos[j] = 0;
}
for (j=0; j<nvec; j++) {
if (vecinosCercanos[j] >= 0)
votos[clases[vecinosCercanos[j]]] ++;
}
votada = 0;
votaciones = votos[0];
for (j=1; j<nClases; j++) {
if (votaciones < votos[j]) {
votaciones = votos[j];
votada = j;
}
}
for (i=0; i<vecinosCercanos.length; i++)
vecinos[i] = vecinosCercanos[i];
return votada;
}
/**
* <p>
* Generates a synthetic example for the minority class from two existing
* examples in the current population
* </p>
*
* @param ra Array with the real values of the first example in the current population
* @param rb Array with the real values of the second example in the current population
* @param na Array with the nominal values of the first example in the current population
* @param nb Array with the nominal values of the second example in the current population
* @param ma Array with the missing values of the first example in the current population
* @param mb Array with the missing values of the second example in the current population
* @param resS Array with the general data about the generated example
* @param resR Array with the real values of the generated example
* @param resN Array with the nominal values of the generated example
* @param resM Array with the missing values of the generated example
*/
void interpolate (double ra[], double rb[], int na[], int nb[], boolean ma[], boolean mb[], double resS[], double resR[], int resN[], boolean resM[]) {
int i;
double diff;
double gap;
int suerte;
for (i=0; i<ra.length; i++) {
if (ma[i] == true && mb[i] == true) {
resM[i] = true;
resS[i] = 0;
} else if (ma[i] == true){
if (entradas[i].getType() == Attribute.REAL) {
resR[i] = rb[i];
resS[i] = (resR[i] + entradas[i].getMinAttribute()) / (entradas[i].getMaxAttribute() - entradas[i].getMinAttribute());
} else if (entradas[i].getType() == Attribute.INTEGER) {
resR[i] = rb[i];
resS[i] = (resR[i] + entradas[i].getMinAttribute()) / (entradas[i].getMaxAttribute() - entradas[i].getMinAttribute());
} else {
resN[i] = nb[i];
resS[i] = (double)resN[i] / (double)(entradas[i].getNominalValuesList().size() - 1);
}
} else if (mb[i] == true) {
if (entradas[i].getType() == Attribute.REAL) {
resR[i] = ra[i];
resS[i] = (resR[i] + entradas[i].getMinAttribute()) / (entradas[i].getMaxAttribute() - entradas[i].getMinAttribute());
} else if (entradas[i].getType() == Attribute.INTEGER) {
resR[i] = ra[i];
resS[i] = (resR[i] + entradas[i].getMinAttribute()) / (entradas[i].getMaxAttribute() - entradas[i].getMinAttribute());
} else {
resN[i] = na[i];
resS[i] = (double)resN[i] / (double)(entradas[i].getNominalValuesList().size() - 1);
}
} else {
resM[i] = false;
if (entradas[i].getType() == Attribute.REAL) {
diff = rb[i] - ra[i];
gap = Randomize.Rand();
resR[i] = ra[i] + gap*diff;
resS[i] = (resR[i] + entradas[i].getMinAttribute()) / (entradas[i].getMaxAttribute() - entradas[i].getMinAttribute());
} else if (entradas[i].getType() == Attribute.INTEGER) {
diff = rb[i] - ra[i];
gap = Randomize.Rand();
resR[i] = Math.round(ra[i] + gap*diff);
resS[i] = (resR[i] + entradas[i].getMinAttribute()) / (entradas[i].getMaxAttribute() - entradas[i].getMinAttribute());
} else {
suerte = Randomize.Randint(0, 2);
if (suerte == 0) {
resN[i] = na[i];
} else {
resN[i] = nb[i];
}
resS[i] = (double)resN[i] / (double)(entradas[i].getNominalValuesList().size() - 1);
}
}
}
}
/**
* <p>
* Obtains the parameters used in the execution of the algorithm and stores
* them in the private variables of the class
* </p>
*
* @param ficheroScript Name of the configuration script that indicates the
* parameters that are going to be used during the execution of the algorithm
*/
public void leerConfiguracion (String ficheroScript) {
String fichero, linea, token;
StringTokenizer lineasFichero, tokens;
byte line[];
int i, j;
ficheroSalida = new String[2];
fichero = Fichero.leeFichero (ficheroScript);
lineasFichero = new StringTokenizer (fichero,"\n\r");
lineasFichero.nextToken();
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
token = tokens.nextToken();
/*Getting the names of the training and test files*/
line = token.getBytes();
for (i=0; line[i]!='\"'; i++);
i++;
for (j=i; line[j]!='\"'; j++);
ficheroTraining = new String (line,i,j-i);
for (i=j+1; line[i]!='\"'; i++);
i++;
for (j=i; line[j]!='\"'; j++);
ficheroTest = new String (line,i,j-i);
/*Getting the path and base name of the results files*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
token = tokens.nextToken();
/*Getting the names of output files*/
line = token.getBytes();
for (i=0; line[i]!='\"'; i++);
i++;
for (j=i; line[j]!='\"'; j++);
ficheroSalida[0] = new String (line,i,j-i);
for (i=j+1; line[i]!='\"'; i++);
i++;
for (j=i; line[j]!='\"'; j++);
ficheroSalida[1] = new String (line,i,j-i);
/*Getting the seed*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
semilla = Long.parseLong(tokens.nextToken().substring(1));
/*Getting the number of neighbors for ENN*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
kENN = Integer.parseInt(tokens.nextToken().substring(1));
/*Getting the number of neighbors*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
kSMOTE = Integer.parseInt(tokens.nextToken().substring(1));
/*Getting the type of SMOTE algorithm*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
token = tokens.nextToken();
token = token.substring(1);
if (token.equalsIgnoreCase("both")) ASMO = 0;
else if (token.equalsIgnoreCase("minority")) ASMO = 1;
else ASMO = 2;
/*Getting the type of balancing in SMOTE*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
token = tokens.nextToken();
token = token.substring(1);
if (token.equalsIgnoreCase("YES")) balance = true;
else balance = false;
/*Getting the quantity of smoting*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
smoting = Double.parseDouble(tokens.nextToken().substring(1));
/*Getting the type of distance function*/
linea = lineasFichero.nextToken();
tokens = new StringTokenizer (linea, "=");
tokens.nextToken();
distanceEu = tokens.nextToken().substring(1).equalsIgnoreCase("Euclidean")?true:false;
}
/**
* This function builds the data matrix for reference data and normalizes inputs values
*/
protected void normalizar () throws CheckException {
int i, j, k;
Instance temp;
double caja[];
StringTokenizer tokens;
boolean nulls[];
/*Check if dataset corresponding with a classification problem*/
if (Attributes.getOutputNumAttributes() < 1) {
throw new CheckException ("This dataset haven?t outputs, so it not corresponding to a classification problem.");
} else if (Attributes.getOutputNumAttributes() > 1) {
throw new CheckException ("This dataset have more of one output.");
}
if (Attributes.getOutputAttribute(0).getType() == Attribute.REAL) {
throw new CheckException ("This dataset have an input attribute with floating values, so it not corresponding to a classification problem.");
}
entradas = Attributes.getInputAttributes();
salida = Attributes.getOutputAttribute(0);
nEntradas = Attributes.getInputNumAttributes();
tokens = new StringTokenizer (training.getHeader()," \n\r");
tokens.nextToken();
relation = tokens.nextToken();
datosTrain = new double[training.getNumInstances()][Attributes.getInputNumAttributes()];
clasesTrain = new int[training.getNumInstances()];
caja = new double[1];
nulosTrain = new boolean[training.getNumInstances()][Attributes.getInputNumAttributes()];
nominalTrain = new int[training.getNumInstances()][Attributes.getInputNumAttributes()];
realTrain = new double[training.getNumInstances()][Attributes.getInputNumAttributes()];
for (i=0; i<training.getNumInstances(); i++) {
temp = training.getInstance(i);
nulls = temp.getInputMissingValues();
datosTrain[i] = training.getInstance(i).getAllInputValues();
for (j=0; j<nulls.length; j++)
if (nulls[j]) {
datosTrain[i][j]=0.0;
nulosTrain[i][j] = true;
}
caja = training.getInstance(i).getAllOutputValues();
clasesTrain[i] = (int) caja[0];
for (k = 0; k < datosTrain[i].length; k++) {
if (Attributes.getInputAttribute(k).getType() == Attribute.NOMINAL) {
nominalTrain[i][k] = (int)datosTrain[i][k];
datosTrain[i][k] /= Attributes.getInputAttribute(k).
getNominalValuesList().size() - 1;
} else {
realTrain[i][k] = datosTrain[i][k];
datosTrain[i][k] -= Attributes.getInputAttribute(k).getMinAttribute();
datosTrain[i][k] /= Attributes.getInputAttribute(k).getMaxAttribute() -
Attributes.getInputAttribute(k).getMinAttribute();
if (Double.isNaN(datosTrain[i][k])){
datosTrain[i][k] = realTrain[i][k];
}
}
}
}
datosTest = new double[test.getNumInstances()][Attributes.getInputNumAttributes()];
clasesTest = new int[test.getNumInstances()];
caja = new double[1];
for (i=0; i<test.getNumInstances(); i++) {
temp = test.getInstance(i);
nulls = temp.getInputMissingValues();
datosTest[i] = test.getInstance(i).getAllInputValues();
for (j=0; j<nulls.length; j++)
if (nulls[j]) {
datosTest[i][j]=0.0;
}
caja = test.getInstance(i).getAllOutputValues();
clasesTest[i] = (int) caja[0];
}
} //end-method
}