yawn.nn.mlp
Class MLPLayer

java.lang.Object
  yawn.nn.Layer
      yawn.nn.mlp.MLPLayer

public class MLPLayer
extends Layer

This a layer is meant to be used in a Multi-Layer Perceptron. It handles the thresholds as an extra weight automaticaly. $Id: MLPLayer.java,v 1.10 2005/04/07 17:28:07 supermarti Exp $

Version:

$Revision: 1.10 $

Author:

Luis Martí (luis dot marti at uc3m dot es)

Field Summary

Fields inherited from class yawn.nn.Layer

adapting, inputSize, nextLayer, units

Constructor Summary

MLPLayer(Layer nextLayer, int inputSize, int nodesSize, java.lang.Class nodesClass)

Method Summary

void	adapt(Pattern deltas, double learningRate, double momentumRate)
Pattern	calculateDeltasAsHiddenLayer(Pattern nextLayerDeltas)
Pattern	calculateDeltasAsOutputLayer(Pattern expectedOutputs)
void	setInput(Pattern input) Does the `threshold as weight' trick.

Methods inherited from class yawn.nn.Layer

connectWith, getActivations, getInputSize, getNodes, isAdapting, output, propagateToNextLayer, reset, setAdapting, setInputSize, size

Methods inherited from class java.lang.Object

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

Constructor Detail

MLPLayer

public MLPLayer(Layer nextLayer,
                int inputSize,
                int nodesSize,
                java.lang.Class nodesClass)

Parameters:

nextLayer -

inputSize -

Method Detail

setInput

public void setInput(Pattern input)

Does the `threshold as weight' trick.

Overrides:

setInput in class Layer

Parameters:

input - the input pattern

calculateDeltasAsOutputLayer

public Pattern calculateDeltasAsOutputLayer(Pattern expectedOutputs)

calculateDeltasAsHiddenLayer

public Pattern calculateDeltasAsHiddenLayer(Pattern nextLayerDeltas)

adapt

public void adapt(Pattern deltas,
                  double learningRate,
                  double momentumRate)