yawn.nn.appart
Class RecognitionLayer

java.lang.Object
  extended by yawn.nn.Layer
      extended by yawn.nn.appart.RecognitionLayer
Direct Known Subclasses:
GasRecognitionLayer

public class RecognitionLayer
extends Layer

A recognition (F2) layer for AppART neural networks. $Id: RecognitionLayer.java,v 1.8 2005/04/07 17:28:25 supermarti Exp $

Version:
$Revision: 1.8 $
Author:
Luis Martí (luis dot marti at uc3m dot es)

Field Summary
protected  GainControlUnitOnMatching g
           
protected  int inputSize
           
protected  Pattern normalizedActivations
           
protected  boolean normDone
           
 
Fields inherited from class yawn.nn.Layer
adapting, nextLayer, units
 
Constructor Summary
RecognitionLayer(Layer nextLayer, GainControlUnitOnMatching aGF2, int inputSize)
           
 
Method Summary
 void calculateNormalizedActivations()
           
 Pattern getNormalizedActivations()
           
 void learn()
           
 void makeNewCategory()
           
 void matching()
           
 void propagateToNextLayer()
          Propagates the normalizedActivations to a prediction layer.
 void reset()
           
 void setGF2(GainControlUnitOnMatching aGF2)
           
 void setInput(Pattern input)
           
 
Methods inherited from class yawn.nn.Layer
connectWith, getActivations, getInputSize, getNodes, isAdapting, output, setAdapting, setInputSize, size
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

g

protected GainControlUnitOnMatching g

inputSize

protected int inputSize

normalizedActivations

protected Pattern normalizedActivations

normDone

protected boolean normDone
Constructor Detail

RecognitionLayer

public RecognitionLayer(Layer nextLayer,
                        GainControlUnitOnMatching aGF2,
                        int inputSize)
Method Detail

calculateNormalizedActivations

public void calculateNormalizedActivations()

getNormalizedActivations

public Pattern getNormalizedActivations()

learn

public void learn()

makeNewCategory

public void makeNewCategory()

matching

public void matching()

propagateToNextLayer

public void propagateToNextLayer()
Propagates the normalizedActivations to a prediction layer.

Overrides:
propagateToNextLayer in class Layer

reset

public void reset()
Overrides:
reset in class Layer
See Also:
Layer.reset()

setGF2

public void setGF2(GainControlUnitOnMatching aGF2)

setInput

public void setInput(Pattern input)
Overrides:
setInput in class Layer
See Also:
Layer.setInput(yawn.util.Pattern)


Copyright © 2003-2005 GIAA, Universidad Carlos III de Madrid. All Rights Reserved.