org.encog.neural.pattern
Class CPNPattern

java.lang.Object
  extended by org.encog.neural.pattern.CPNPattern
All Implemented Interfaces:
NeuralNetworkPattern

public class CPNPattern
extends Object
implements NeuralNetworkPattern

Pattern that creates a CPN neural network.


Field Summary
static String TAG_INSTAR
          The tag for the INSTAR layer.
static String TAG_OUTSTAR
          The tag for the OUTSTAR layer.
 
Constructor Summary
CPNPattern()
           
 
Method Summary
 void addHiddenLayer(int count)
          Not used, will throw an error.
 void clear()
          Clear any parameters that were set.
 MLMethod generate()
          Generate the network.
 void setActivationFunction(ActivationFunction activation)
          This method will throw an error.
 void setInputNeurons(int count)
          Set the number of input neurons.
 void setInstarCount(int instarCount)
          Set the number of neurons in the instar layer.
 void setOutputNeurons(int count)
          Set the number of output neurons.
 void setOutstarCount(int outstarCount)
          Set the number of neurons in the outstar level, this level is mapped to the "output" level.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

TAG_INSTAR

public static final String TAG_INSTAR
The tag for the INSTAR layer.

See Also:
Constant Field Values

TAG_OUTSTAR

public static final String TAG_OUTSTAR
The tag for the OUTSTAR layer.

See Also:
Constant Field Values
Constructor Detail

CPNPattern

public CPNPattern()
Method Detail

addHiddenLayer

public void addHiddenLayer(int count)
Not used, will throw an error. CPN networks already have a predefined hidden layer called the instar layer.

Specified by:
addHiddenLayer in interface NeuralNetworkPattern
Parameters:
count - NOT USED

clear

public void clear()
Clear any parameters that were set.

Specified by:
clear in interface NeuralNetworkPattern

generate

public MLMethod generate()
Generate the network.

Specified by:
generate in interface NeuralNetworkPattern
Returns:
The generated network.

setActivationFunction

public void setActivationFunction(ActivationFunction activation)
This method will throw an error. The CPN network uses predefined activation functions.

Specified by:
setActivationFunction in interface NeuralNetworkPattern
Parameters:
activation - NOT USED

setInputNeurons

public void setInputNeurons(int count)
Set the number of input neurons.

Specified by:
setInputNeurons in interface NeuralNetworkPattern
Parameters:
count - The input neuron count.

setInstarCount

public void setInstarCount(int instarCount)
Set the number of neurons in the instar layer. This level is essentially a hidden layer.

Parameters:
instarCount - The instar count.

setOutputNeurons

public void setOutputNeurons(int count)
Set the number of output neurons. Calling this method maps to setting the number of neurons in the outstar layer.

Specified by:
setOutputNeurons in interface NeuralNetworkPattern
Parameters:
count - The count.

setOutstarCount

public void setOutstarCount(int outstarCount)
Set the number of neurons in the outstar level, this level is mapped to the "output" level.

Parameters:
outstarCount - The outstar count.


Copyright © 2014. All Rights Reserved.