org.encog.neural.networks.logic
Class SimpleRecurrentLogic

java.lang.Object
  org.encog.neural.networks.logic.FeedforwardLogic
      org.encog.neural.networks.logic.SimpleRecurrentLogic

All Implemented Interfaces:

java.io.Serializable, NeuralLogic

Direct Known Subclasses:

SOMLogic, ThermalLogic

public class SimpleRecurrentLogic
extends FeedforwardLogic

Provides the neural logic for an Simple Recurrent Network (SRN) type network. This class is used for the Elman and Jordan networks. This class will work just fine for a feedforward neural network, however it is not efficient.

See Also:

Serialized Form

Constructor Summary

SimpleRecurrentLogic()

Method Summary

void	preprocessLayer(Layer layer, NeuralData input, Synapse source) Handle recurrent layers.

Methods inherited from class org.encog.neural.networks.logic.FeedforwardLogic

compute, getNetwork, init

Methods inherited from class java.lang.Object

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

Constructor Detail

SimpleRecurrentLogic

public SimpleRecurrentLogic()

Method Detail

preprocessLayer

public void preprocessLayer(Layer layer,
                            NeuralData input,
                            Synapse source)

Handle recurrent layers. See if there are any recurrent layers before the specified layer that must affect the input.

Overrides:

preprocessLayer in class FeedforwardLogic

Parameters:

layer - The layer being processed, see if there are any recurrent connections to this.

input - The input to the layer, will be modified with the result from any recurrent layers.

source - The source synapse.