keel.Algorithms.Genetic_Rule_Learning.Ant_Miner_Plus

Class ACO

java.lang.Object

keel.Algorithms.Genetic_Rule_Learning.Ant_Miner_Plus.ACO

public class ACO
extends java.lang.Object

Title: Ant Colony Optimization

Description: Classification Algorithm by ACO (Advanced Ant Miner +).

Version:

1.0

Author:

Vicente Rubén del Pino Ruiz

Constructor Summary

Constructors

Constructor and Description
ACO() Default constructor.
ACO(java.lang.String fTrainPrep, java.lang.String fTrain, java.lang.String fTestOriginal, java.lang.String fStop, java.lang.String fSalidaTrain, java.lang.String fSalidaTest, java.lang.String fSalidaResult, int nHormigas, int maxDatos, int minCasos, int maxIter, float minimoFeromona, float maximoFeromona, float alfaIn, float betaIn, long semillaOriginal) Parameter constructor.

Method Summary

Modifier and Type	Method and Description
void	inicializaSiTieneCondicionVacia() Initializes a vector that indicates if the corresponding attribute of the rule can be assigned with a null/empty condition.
void	muestraResultados() Prints on the standard output the algorithm results.
boolean	OK() All is OK is the data-sets have not got any continuos values
void	run() Executes the algorithm.
void	sacaResultadosAFicheros() 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

ACO

public ACO()

Default constructor. Nothing is done.

ACO

public ACO(java.lang.String fTrainPrep,
           java.lang.String fTrain,
           java.lang.String fTestOriginal,
           java.lang.String fStop,
           java.lang.String fSalidaTrain,
           java.lang.String fSalidaTest,
           java.lang.String fSalidaResult,
           int nHormigas,
           int maxDatos,
           int minCasos,
           int maxIter,
           float minimoFeromona,
           float maximoFeromona,
           float alfaIn,
           float betaIn,
           long semillaOriginal)

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.

fStop - String Stop dataset filename.

fSalidaTrain - String Training output filename.

fSalidaTest - String Test output filename.

fSalidaResult - String Global results filename.

nHormigas - int Number of ants considered.

maxDatos - int Maximum number of uncovered data.

minCasos - int Minimum number of cases that a rule must cover.

maxIter - int Maximum iterations.

minimoFeromona - float Minimum pheromone value.

maximoFeromona - float Maximum pheromone value.

alfaIn - float alpha parameter.

betaIn - float beta parameter.

semillaOriginal - long Seed for the random numbers generator.

Method Detail

OK

public boolean OK()

All is OK is the data-sets have not got any continuos values

Returns:

boolean a flag for the existence of continous values in the data-set

inicializaSiTieneCondicionVacia

public void inicializaSiTieneCondicionVacia()

Initializes a vector that indicates if the corresponding attribute of the rule can be assigned with a null/empty condition.

run

public void run()

Executes the algorithm.

sacaResultadosAFicheros

public void sacaResultadosAFicheros()

Prints all the results on the output results files.

muestraResultados

public void muestraResultados()

Prints on the standard output the algorithm results. Prints each extrated rule followed by its acurracies for training and test.