keel.Algorithms.Genetic_Rule_Learning.PSO_ACO

Class PsoAco

java.lang.Object

keel.Algorithms.Genetic_Rule_Learning.PSO_ACO.PsoAco

public class PsoAco
extends java.lang.Object

Title: PsoAco (Particle Swarm Optimization and Ant Colony Optimization)

Description: Hybridization of the PSO and ACO algorithms to conform a rules based classification algorithm.

Version:

1.0

Author:

Vicente Rubén del Pino Ruiz

Constructor Summary

Constructors

Constructor and Description
PsoAco() Default constructor.
PsoAco(java.lang.String fTrainPrep, java.lang.String fTrain, java.lang.String fTestOriginal, java.lang.String fSalidaTrain, java.lang.String fSalidaTest, java.lang.String fSalidaResult, long semillaOriginal, int maxCasosSinCubrirO, int numParticulasO, int tamEntornoO, int maxIteracionesO, int minimoCasosCubiertosO, float xO, float c1O, float c2O, int banderaO) Parameter constructor.

Method Summary

Modifier and Type	Method and Description
Regla	creaReglaGenerica() Creates a default rule en case that any of the generated ones by the algorithm can be applied.
void	run() Executes the algorithm.
void	sacaResultadosAFicheros(java.util.Vector reglasDescubiertas) Prints all the results on the output results files.

Methods inherited from class java.lang.Object

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

Constructor Detail

PsoAco

public PsoAco()

Default constructor. Nothing is done.

PsoAco

public PsoAco(java.lang.String fTrainPrep,
              java.lang.String fTrain,
              java.lang.String fTestOriginal,
              java.lang.String fSalidaTrain,
              java.lang.String fSalidaTest,
              java.lang.String fSalidaResult,
              long semillaOriginal,
              int maxCasosSinCubrirO,
              int numParticulasO,
              int tamEntornoO,
              int maxIteracionesO,
              int minimoCasosCubiertosO,
              float xO,
              float c1O,
              float c2O,
              int banderaO)

Parameter constructor. Build the algorithm object with the parameter given.

Parameters:

fTrainPrep - String Preprocessed Training filename.

fTrain - String Full training filename.

fTestOriginal - String Test filename.

fSalidaTrain - String Training output filename.

fSalidaTest - String Test output filename.

fSalidaResult - String Global results filename.

maxCasosSinCubrirO - maximum number of uncovered examples allowed.

numParticulasO - Number of particles considered.

tamEntornoO - Environment size.

maxIteracionesO - Maximum iterations.

minimoCasosCubiertosO - Minimum number of cases that a rule must cover.

xO - X

c1O - Coefficient c1.

c2O - Coefficient cw.

banderaO - flag for the election of the condition type.

semillaOriginal - long Seed for the random numbers generator.

Method Detail

creaReglaGenerica

public Regla creaReglaGenerica()

Creates a default rule en case that any of the generated ones by the algorithm can be applied.

Returns:

Regla a default rule: assign the same class to every example.

run

public void run()

Executes the algorithm.

sacaResultadosAFicheros

public void sacaResultadosAFicheros(java.util.Vector reglasDescubiertas)

Prints all the results on the output results files.

Parameters:

reglasDescubiertas - extracted rules vector.