keel.Algorithms.Instance_Generation.ENPC

Class ENPCGenerator

java.lang.Object

keel.Algorithms.Instance_Generation.Basic.PrototypeGenerator

keel.Algorithms.Instance_Generation.ENPC.ENPCGenerator

public class ENPCGenerator
extends PrototypeGenerator

ENPC prototype generator.

Version:

1.0

Author:

Isaac Triguero

Field Summary

Fields

Modifier and Type	Field and Description
protected int	numberOfClass Class number.
protected int	numberOfPrototypes Number of prototypes.

Fields inherited from class keel.Algorithms.Instance_Generation.Basic.PrototypeGenerator

algorithmName, generatedDataSet, SEED, seedDefaultValueList, trainingDataSet

Constructor Summary

Constructors

Constructor and Description
ENPCGenerator(PrototypeSet _trainingDataSet, int k, int max) Build a new ENPCGenerator Algorithm
ENPCGenerator(PrototypeSet t, Parameters parameters) Build a new RSPGenerator Algorithm

Method Summary

Modifier and Type	Method and Description
protected PrototypeSet	die(PrototypeSet classifier, double[] quality) Die operator.
protected double	expectation(double sj, PrototypeSet S) Function expectation of class sj in the S prototypeSet., see equation (3).
protected void	fight(PrototypeSet classifier, PrototypeSet[][] V, double[] quality)
static void	main(java.lang.String[] args) General main for all the prototoype generators Arguments: 0: Filename with the training data set to be condensed. 1: Filename which contains the test data set. 3: Seed of the random number generator.
protected void	move(PrototypeSet classifier, PrototypeSet[][] V) Moves all to the centroid.
protected void	mutation(PrototypeSet classifier, PrototypeSet[][] V) Mutation operator.
PrototypeSet	reduceSet() Generate a reduced prototype set by the ENPCGenerator method.
protected int	regions(double sj, PrototypeSet S) Function regions, see equation (2).
protected PrototypeSet[][]	reproduction(PrototypeSet classifier, PrototypeSet[][] V) Reproduction operator.

Methods inherited from class keel.Algorithms.Instance_Generation.Basic.PrototypeGenerator

absoluteAccuracy, absoluteAccuracyAndError, absoluteAccuracyKNN, accuracy, accuracy2, desordenar_vector_sin, desordenar_vector, execute, generateReducedDataSet, getResultingAccuracy, getResultingAccuracy, getResults, getResultsOfAccuracy, getResultsOfAccuracy, getSeed, getSetSizeFromPercentage, getSetSizeFromPercentage, getTime, inic_vector_sin, inic_vector, saveResultsOfAccuracyIn, saveResultsOfAccuracyIn, selecRandomSet, setSeed, showResultsOfAccuracy, showResultsOfAccuracy

Methods inherited from class java.lang.Object

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

Field Detail

numberOfPrototypes

protected int numberOfPrototypes

Number of prototypes.

numberOfClass

protected int numberOfClass

Class number.

Constructor Detail

ENPCGenerator

public ENPCGenerator(PrototypeSet _trainingDataSet,
                     int k,
                     int max)

Build a new ENPCGenerator Algorithm

Parameters:

_trainingDataSet - training dataset.

k - k parameter.

max - maximum number of iterations

ENPCGenerator

public ENPCGenerator(PrototypeSet t,
                     Parameters parameters)

Build a new RSPGenerator Algorithm

Parameters:

t - Original prototype set to be reduced.

parameters - Parameters of the algorithm (only % of reduced set).

Method Detail

regions

protected int regions(double sj,
                      PrototypeSet S)

Function regions, see equation (2).

Parameters:

sj -

S -

Returns:

expectation

protected double expectation(double sj,
                             PrototypeSet S)

Function expectation of class sj in the S prototypeSet., see equation (3).

Parameters:

sj -

S -

Returns:

mutation

protected void mutation(PrototypeSet classifier,
                        PrototypeSet[][] V)

Mutation operator. Label each prototype with the most populate class in each region. It is the way of obtain the main class that is usually use when unsupervised learning is applied to supervise classification. But not onl in a posteriori phase. Remember tha the quality of each prototype depends on the relationship among the number of patterns in its regions and....,

Parameters:

classifier - classifier prototype set

V - prototype set population.

reproduction

protected PrototypeSet[][] reproduction(PrototypeSet classifier,
                                        PrototypeSet[][] V)

Reproduction operator. Introduce new prototypes in the classifier, each prototype has the opportunity to introduce a new prototype in order to increase its own quality. If a region r1 has two non-empty sets V11 and V12 we need to have another region r2 that contain V12

Parameters:

classifier - classifier prototype set

V - prototype set population.

Returns:

new prototype set population.

fight

protected void fight(PrototypeSet classifier,
                     PrototypeSet[][] V,
                     double[] quality)

move

protected void move(PrototypeSet classifier,
                    PrototypeSet[][] V)

Moves all to the centroid.

Parameters:

classifier - classifier prototype set

V - prototype set population.

die

protected PrototypeSet die(PrototypeSet classifier,
                           double[] quality)

Die operator. Removes all the prototypes that do not have a good quality.

Parameters:

classifier - classifier prototype set

quality - quality of the prototypes in the set.

Returns:

Modified classifier prototype set.

reduceSet

public PrototypeSet reduceSet()

Generate a reduced prototype set by the ENPCGenerator method.

Overrides:

reduceSet in class PrototypeGenerator

Returns:

Reduced set by ENPCGenerator's method.

main

public static void main(java.lang.String[] args)

General main for all the prototoype generators Arguments: 0: Filename with the training data set to be condensed. 1: Filename which contains the test data set. 3: Seed of the random number generator. Always. ************************** 4: .Number of blocks

Parameters:

args - Arguments of the main function.