keel.Algorithms.Rule_Learning.Riona

Class Riona

java.lang.Object

keel.Algorithms.Rule_Learning.Riona.Riona

public class Riona
extends java.lang.Object

Main procedures of Rionasd algorithm

Since:

JDK1.2

Version:

1.2

Author:

Written by Rosa Venzala 19/09/2008, Modified by Xavi Solé (La Salle, Ramón Llull University - Barcelona) 16/12/2008

Constructor Summary

Constructors

Constructor and Description
Riona() Riona default constructor
Riona(java.lang.String trainFile, java.lang.String testFile, java.lang.String trainOutFile, java.lang.String testOutFile, java.lang.String fOut, long lSeed) Riona constructor

Method Summary

Modifier and Type	Method and Description
Complex	createRuleTestTrain(double[] tst, int numItst, double[] trn, int numItrn, int clase, int nClasses, boolean es_train) Creates a local rule
int	findOptimalK(int kmax) Calculates the optimum size of the neighborhood for the training set
boolean	isConsistent(Complex R, int[] verifySet, int util) Inidcates if a rule is consistent with a determined set of examples

Methods inherited from class java.lang.Object

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

Constructor Detail

Riona

public Riona()

Riona default constructor

Riona

public Riona(java.lang.String trainFile,
             java.lang.String testFile,
             java.lang.String trainOutFile,
             java.lang.String testOutFile,
             java.lang.String fOut,
             long lSeed)

Riona constructor

Parameters:

trainFile - Training dataset file.

testOutFile - Test output file.

testFile - Test dataset file.

trainOutFile - Traininf output file.

fOut - Output file.

lSeed - Seed for random purpose.

Method Detail

createRuleTestTrain

public Complex createRuleTestTrain(double[] tst,
                                   int numItst,
                                   double[] trn,
                                   int numItrn,
                                   int clase,
                                   int nClasses,
                                   boolean es_train)

Creates a local rule

Parameters:

tst - the test example

numItst - the position of the example in the set

trn - the train example

numItrn - the position of the example in the set

clase - the consequent of the rule

nClasses - the total number of classes

es_train - identify if it is a training process.

Returns:

a local rule

isConsistent

public boolean isConsistent(Complex R,
                            int[] verifySet,
                            int util)

Inidcates if a rule is consistent with a determined set of examples

Parameters:

R - rule to check.

verifySet - set of examples.

util - size of the set.

Returns:

true if is consistent

findOptimalK

public int findOptimalK(int kmax)

Calculates the optimum size of the neighborhood for the training set

Parameters:

kmax - max malue for neighborhood size

Returns: