org.encog.neural.networks.training.propagation
Class PropagationSynapse

java.lang.Object
  org.encog.neural.networks.training.propagation.PropagationSynapse

public class PropagationSynapse
extends java.lang.Object

The back propagation training algorithms requires training data to be stored for each of the synapses. The propagation class creates a PropagationSynapse object for each of the synapses in the neural network that it is training.

Constructor Summary

PropagationSynapse(Synapse synapse)
Construct a PropagationSynapse object that corresponds to a specific synapse.

Method Summary

Matrix	getAccMatrixGradients()
Matrix	getDeltas()
Matrix	getLastMatrixGradients()
Synapse	getSynapse()
void	setLastMatrixGradients(Matrix d) Set the matrix gradients.
java.lang.String	toString()

Methods inherited from class java.lang.Object

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

Constructor Detail

PropagationSynapse

public PropagationSynapse(Synapse synapse)

Construct a PropagationSynapse object that corresponds to a specific synapse.

Parameters:

synapse - The back propagation training object.

Method Detail

getAccMatrixGradients

public Matrix getAccMatrixGradients()

Returns:

The accumulated matrix gradients.

getDeltas

public Matrix getDeltas()

Returns:

The matrix deltas, these changes are applied to the matrix in some propagation techniques.

getLastMatrixGradients

public Matrix getLastMatrixGradients()

Returns:

The matrix gradients from the pervious iteration.

getSynapse

public Synapse getSynapse()

Returns:

Get the synapse that this object is linked with.

setLastMatrixGradients

public void setLastMatrixGradients(Matrix d)

Set the matrix gradients.

Parameters:

d - The matrix to set.

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object

Returns:

This object as a string.