com.huguesjohnson.jail.neuralNetwork
Class BackPropagationLayer

java.lang.Object
  |
  +--com.huguesjohnson.jail.neuralNetwork.Layer
        |
        +--com.huguesjohnson.jail.neuralNetwork.BackPropagationLayer
All Implemented Interfaces:
java.lang.Runnable, java.io.Serializable

public class BackPropagationLayer
extends Layer

class BackPropagationLayer

See Also:
Serialized Form

Constructor Summary
BackPropagationLayer()
          default constructor
BackPropagationLayer(int layerSize, int inputSize, double punishRate, double rewardRate, double errorThreshold, double minWeight, double maxWeight)
          constructor
 
Method Summary
 void computeOutput()
          computes the output for the layer
 void runComplete()
          method invoked when run is complete
 void train()
          move actual output closer to desired output by default find neuron with greatest error and adjust it
 
Methods inherited from class com.huguesjohnson.jail.neuralNetwork.Layer
adjustTrainingRate, computeError, getDesiredOutput, getDoneTraining, getError, getNeuron, getOutput, getOutput, getSize, initOutput, insertNeuron, insertNeuron, removeNeuron, run, setDesiredOutput, setDoneTraining, setInput, setNeuron, setOutput, start, stop, toString
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait
 

Constructor Detail

BackPropagationLayer

public BackPropagationLayer()
default constructor


BackPropagationLayer

public BackPropagationLayer(int layerSize,
                            int inputSize,
                            double punishRate,
                            double rewardRate,
                            double errorThreshold,
                            double minWeight,
                            double maxWeight)
constructor

Parameters:
layerSize - the number of neurons in the layer
inputSize - the number of inputs to the layer
punishRate - the punish rate for every neuron in the layer
rewardRate - the reward rate for every neuron in the layer
errorThreshold - the error threshold for every neuron in the layer
minWeight - minimum weight on a neuron link
maxWeight - maximum weight on a neuron link
Method Detail

runComplete

public void runComplete()
method invoked when run is complete

Overrides:
runComplete in class Layer

train

public void train()
move actual output closer to desired output by default find neuron with greatest error and adjust it

Specified by:
train in class Layer

computeOutput

public void computeOutput()
computes the output for the layer

Specified by:
computeOutput in class Layer