edu.udo.cs.yale.operator.features.construction
Class UnbalancedCrossover

java.lang.Object
  edu.udo.cs.yale.operator.features.selection.SelectionCrossover
      edu.udo.cs.yale.operator.features.construction.UnbalancedCrossover

All Implemented Interfaces:

PopulationOperator

public class UnbalancedCrossover
extends SelectionCrossover

This PopulationOperator applies a crossover on two example sets. Crossover type can be ONE_POINT, UNIFORM, or SHUFFLE. In difference to SelectionCrossover the attribute vectors can have different lengths.
This crossover type should only be used for SINGLE_VALUEs, i.e. attributes without a block number (blocknumber can and should be assigned to value series attributes)!

Version:

$Id: UnbalancedCrossover.java,v 1.1 2006/04/14 07:47:17 ingomierswa Exp $

Author:

Ingo Mierswa

Nested Class Summary

private class

UnbalancedCrossover.AttributeWeightContainer

Field Summary

private java.util.Random

random

Fields inherited from class edu.udo.cs.yale.operator.features.selection.SelectionCrossover

CROSSOVER_TYPES, ONE_POINT, SHUFFLE, UNIFORM

Constructor Summary

UnbalancedCrossover(int type, double prob, java.util.Random random)
Creates a new generating crossover with the given type which will be applied with the given probability.

Method Summary

private void	addAttributeToSwapList(AttributeWeightedExampleSet exampleSet, int index, java.util.List<UnbalancedCrossover.AttributeWeightContainer> dummyList)
void	crossover(AttributeWeightedExampleSet es1, AttributeWeightedExampleSet es2) Applies the crossover.
private void	mergeAttributes(AttributeWeightedExampleSet exampleSet, java.util.List<UnbalancedCrossover.AttributeWeightContainer> attributeWeights)

Methods inherited from class edu.udo.cs.yale.operator.features.selection.SelectionCrossover

getType, operate, performOperation

Methods inherited from class java.lang.Object

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

Field Detail

random

private java.util.Random random

Constructor Detail

UnbalancedCrossover

public UnbalancedCrossover(int type,
                           double prob,
                           java.util.Random random)

Creates a new generating crossover with the given type which will be applied with the given probability.

Method Detail

crossover

public void crossover(AttributeWeightedExampleSet es1,
                      AttributeWeightedExampleSet es2)

Applies the crossover. Works directly on the given example sets.

Overrides:

crossover in class SelectionCrossover

addAttributeToSwapList

private void addAttributeToSwapList(AttributeWeightedExampleSet exampleSet,
                                    int index,
                                    java.util.List<UnbalancedCrossover.AttributeWeightContainer> dummyList)

mergeAttributes

private void mergeAttributes(AttributeWeightedExampleSet exampleSet,
                             java.util.List<UnbalancedCrossover.AttributeWeightContainer> attributeWeights)