com.rapidminer.tools.math.optimization.ec.es
Class StochasticUniversalSampling

java.lang.Object
  com.rapidminer.tools.math.optimization.ec.es.StochasticUniversalSampling

All Implemented Interfaces:

PopulationOperator

public class StochasticUniversalSampling
extends java.lang.Object
implements PopulationOperator

Similar to a the roulette wheel selection the fitness values of all individuals build a partition of the 360 degrees of a wheel. The wheel is turned only once and the individuals are selected based on equidistant marks on the wheel. Optionally the best individual is also kept.

Author:

Ingo Mierswa ingomierswa Exp $

Constructor Summary

StochasticUniversalSampling(int popSize, boolean keepBest, java.util.Random random)

Method Summary

double	filterFitness(double fitness) Subclasses may override this method and recalculate the fitness based on the given one, e.g.
void	operate(Population population)
boolean	performOperation(int generation) The default implementation returns true for every generation.

Methods inherited from class java.lang.Object

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

Constructor Detail

StochasticUniversalSampling

public StochasticUniversalSampling(int popSize,
                                   boolean keepBest,
                                   java.util.Random random)

Method Detail

performOperation

public boolean performOperation(int generation)

The default implementation returns true for every generation.

filterFitness

public double filterFitness(double fitness)

Subclasses may override this method and recalculate the fitness based on the given one, e.g. Boltzmann selection or scaled selection. The default implementation simply returns the given fitness.

operate

public void operate(Population population)

Specified by:

operate in interface PopulationOperator