keel.Algorithms.Decision_Trees.DT_GA

Class Individuo

java.lang.Object

keel.Algorithms.Decision_Trees.DT_GA.Individuo

All Implemented Interfaces:

java.lang.Comparable

public class Individuo
extends java.lang.Object
implements java.lang.Comparable

Title: Individuo (Individual).

Description: This class implements an individual for the genetic algorithm, used to build the decision trees.

Version:

1.0

Author:

not attributable

Constructor Summary

Constructors

Constructor and Description
Individuo() Default Constructor.
Individuo(boolean[] antecedentes, java.lang.String clase, myDataset train, int codigo) Paramenter constructor.
Individuo(Individuo padre, Individuo madre, int puntoCorte) Create a new individual by crossing other two at the position given.
Individuo(myDataset train, int pos_ejemplo) Paramenter constructor.

Method Summary

Modifier and Type	Method and Description
double	clasifica(int[] ejemplos, int nEjemplos) Computes and returns the fitness of the individual as the product of sensitivity and specificity of the given examples classification.
double	clasificaLarge(int[] ejemplos, int nEjemplos) Computes and returns the fitness of the individual as the product of sensitivity and specificity of the given examples classification.
Individuo	clone()
int	compareTo(java.lang.Object a)
Regla	convertir() Converts the chromosome (individual) representation to a valid rule of the decision tree.
void	mutar(double mutProb) Mutates the individual with the given probability.
java.lang.String	printString() Returns a String representation of the Individual.
void	pruning(double[] norm_acc) Prune the individual using accuracy given as parameter.
void	pruning(double infoG, int nEjemplos, int[] ejemplos) Prune the individual using the information gain and the examples passed as argument.
int	size() Returns the size of the chromosome.

Methods inherited from class java.lang.Object

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

Constructor Detail

Individuo

public Individuo()

Default Constructor. Basic structures will be initialized.

Individuo

public Individuo(boolean[] antecedentes,
                 java.lang.String clase,
                 myDataset train,
                 int codigo)

Paramenter constructor. The individual structures will be initialized with the parameters given.

Parameters:

antecedentes - Selected antecedents (attributes).

clase - Class for the individual.

train - Training dataset.

codigo - Individual's code.

Individuo

public Individuo(myDataset train,
                 int pos_ejemplo)

Paramenter constructor. The individual structures will be initialized with the parameters given.

Parameters:

train - Training dataset.

pos_ejemplo - Example used to build randomly the antecedents vector.

Individuo

public Individuo(Individuo padre,
                 Individuo madre,
                 int puntoCorte)

Create a new individual by crossing other two at the position given.

Parameters:

padre - First individual.

madre - Second individual.

puntoCorte - Cross point.

Method Detail

printString

public java.lang.String printString()

Returns a String representation of the Individual.

Returns:

a String representation of the Individual.

convertir

public Regla convertir()

Converts the chromosome (individual) representation to a valid rule of the decision tree.

Returns:

the corresponding rule coded on the individual.

clone

public Individuo clone()

Overrides:

clone in class java.lang.Object

clasifica

public double clasifica(int[] ejemplos,
                        int nEjemplos)

Computes and returns the fitness of the individual as the product of sensitivity and specificity of the given examples classification.

Parameters:

ejemplos - examples indeces to be considered.

nEjemplos - number of examples.

Returns:

the computed fitness.

clasificaLarge

public double clasificaLarge(int[] ejemplos,
                             int nEjemplos)

Computes and returns the fitness of the individual as the product of sensitivity and specificity of the given examples classification.

Parameters:

ejemplos - examples indeces to be considered.

nEjemplos - number of examples.

Returns:

the computed fitness.

size

public int size()

Returns the size of the chromosome.

Returns:

the size of the chromosome.

mutar

public void mutar(double mutProb)

Mutates the individual with the given probability.

Parameters:

mutProb - mutation probability.

pruning

public void pruning(double infoG,
                    int nEjemplos,
                    int[] ejemplos)

Prune the individual using the information gain and the examples passed as argument.

Parameters:

infoG - Information gain given.

nEjemplos - Number of examples considered.

ejemplos - Training examples considered.

pruning

public void pruning(double[] norm_acc)

Prune the individual using accuracy given as parameter.

Parameters:

norm_acc - Accuracies for each antedecent.

compareTo

public int compareTo(java.lang.Object a)

Specified by:

compareTo in interface java.lang.Comparable