org.encog.neural.neat.training.opp.links

Class MutatePerturbLinkWeight

java.lang.Object

org.encog.neural.neat.training.opp.links.MutatePerturbLinkWeight

All Implemented Interfaces:

java.io.Serializable, MutateLinkWeight

public class MutatePerturbLinkWeight
extends java.lang.Object
implements MutateLinkWeight, java.io.Serializable

Mutate weight links by perturbing their weights. This will be done by adding a Gaussian random number with the specified sigma. The sigma specifies the standard deviation of the random number. Because the random numbers are clustered at zero, this can be either an increase or decrease. ----------------------------------------------------------------------------- http://www.cs.ucf.edu/~kstanley/ Encog's NEAT implementation was drawn from the following three Journal Articles. For more complete BibTeX sources, see NEATNetwork.java. Evolving Neural Networks Through Augmenting Topologies Generating Large-Scale Neural Networks Through Discovering Geometric Regularities Automatic feature selection in neuroevolution

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
MutatePerturbLinkWeight(double theSigma) Construct the perturbing mutator.

Method Summary

Modifier and Type	Method and Description
EvolutionaryAlgorithm	getTrainer()
void	init(EvolutionaryAlgorithm theTrainer) Setup the link mutator.
void	mutateWeight(java.util.Random rnd, NEATLinkGene linkGene, double weightRange) Perform the weight mutation on the specified link.
java.lang.String	toString()

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Detail

MutatePerturbLinkWeight

public MutatePerturbLinkWeight(double theSigma)

Construct the perturbing mutator.

Parameters:

theSigma - The sigma (standard deviation) for all random numbers.

Method Detail

getTrainer

public EvolutionaryAlgorithm getTrainer()

Specified by:

getTrainer in interface MutateLinkWeight

Returns:

The training class that this mutator is being used with.

init

public void init(EvolutionaryAlgorithm theTrainer)

Setup the link mutator.

Specified by:

init in interface MutateLinkWeight

Parameters:

theTrainer - The training class that this mutator is used with.

mutateWeight

public void mutateWeight(java.util.Random rnd,
                         NEATLinkGene linkGene,
                         double weightRange)

Perform the weight mutation on the specified link.

Specified by:

mutateWeight in interface MutateLinkWeight

Parameters:

rnd - A random number generator.

linkGene - The link to mutate.

weightRange - The weight range, weights are between -weightRange and +weightRange.

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object