org.encog.neural.neat.training
Class NEATInnovationList

java.lang.Object
  org.encog.ml.genetic.innovation.BasicInnovationList
      org.encog.neural.neat.training.NEATInnovationList

All Implemented Interfaces:

Serializable, InnovationList

public class NEATInnovationList
extends BasicInnovationList
implements Serializable

Implements a NEAT innovation list. NeuroEvolution of Augmenting Topologies (NEAT) is a genetic algorithm for the generation of evolving artificial neural networks. It was developed by Ken Stanley while at The University of Texas at Austin. http://www.cs.ucf.edu/~kstanley/

See Also:

Serialized Form

Constructor Summary

NEATInnovationList()
The default constructor, used mainly for persistance.

NEATInnovationList(Population population, Chromosome links, Chromosome neurons)
Construct an innovation list.

Method Summary

NEATInnovation	checkInnovation(long in, long out, NEATInnovationType type) Check to see if we already have an innovation.
NEATNeuronGene	createNeuronFromID(long neuronID) Create a new neuron gene from an id.
void	createNewInnovation(long in, long out, NEATInnovationType type) Create a new innovation.
long	createNewInnovation(long from, long to, NEATInnovationType innovationType, NEATNeuronType neuronType, double x, double y) Create a new innovation.
void	setPopulation(NEATPopulation population)

Methods inherited from class org.encog.ml.genetic.innovation.BasicInnovationList

add, get, getInnovations

Methods inherited from class java.lang.Object

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

Constructor Detail

NEATInnovationList

public NEATInnovationList()

The default constructor, used mainly for persistance.

NEATInnovationList

public NEATInnovationList(Population population,
                          Chromosome links,
                          Chromosome neurons)

Construct an innovation list.

Parameters:

population - The population.

links - The links.

neurons - THe neurons.

Method Detail

checkInnovation

public NEATInnovation checkInnovation(long in,
                                      long out,
                                      NEATInnovationType type)

Check to see if we already have an innovation.

Parameters:

in - The input neuron.

out - THe output neuron.

type - The type.

Returns:

The innovation, either new or existing if found.

createNeuronFromID

public NEATNeuronGene createNeuronFromID(long neuronID)

Create a new neuron gene from an id.

Parameters:

neuronID - The neuron id.

Returns:

The neuron gene.

createNewInnovation

public void createNewInnovation(long in,
                                long out,
                                NEATInnovationType type)

Create a new innovation.

Parameters:

in - The input neuron.

out - The output neuron.

type - The type.

createNewInnovation

public long createNewInnovation(long from,
                                long to,
                                NEATInnovationType innovationType,
                                NEATNeuronType neuronType,
                                double x,
                                double y)

Create a new innovation.

Parameters:

from - The from neuron.

to - The to neuron.

innovationType - THe innovation type.

neuronType - The neuron type.

x - The x-coordinate.

y - The y-coordinate.

Returns:

The new innovation.

setPopulation

public void setPopulation(NEATPopulation population)