Package ciir.umass.edu.learning.neuralnet

Class Layer

java.lang.Object

ciir.umass.edu.learning.neuralnet.Layer

public class Layer
extends java.lang.Object

Author:

vdang This class implements layers of neurons in neural networks.

Field Summary

Fields

Modifier and Type	Field	Description
protected java.util.List<Neuron>	neurons

Constructor Summary

Constructors

Constructor	Description
Layer(int size)
Layer(int size, int nType)

Method Summary

Modifier and Type	Method	Description
void	clearOutputs()
void	computeDelta(PropParameter param)	[Only for output layers] Compute delta for all neurons in the this (output) layer
void	computeOutput()
void	computeOutput(int i)	Have all neurons in this layer compute its output
Neuron	get(int k)
int	size()
void	updateDelta(PropParameter param)	Update delta from neurons in the previous layers
void	updateWeight(PropParameter param)

Methods inherited from class java.lang.Object

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

Field Detail

neurons

protected java.util.List<Neuron> neurons

Constructor Detail

Layer

public Layer(int size)

Layer

public Layer(int size,
             int nType)

Parameters:

size -

nType - 0 for pair; 1 for list.

Method Detail

get

public Neuron get(int k)

size

public int size()

computeOutput

public void computeOutput(int i)

Have all neurons in this layer compute its output

computeOutput

public void computeOutput()

clearOutputs

public void clearOutputs()

computeDelta

public void computeDelta(PropParameter param)

[Only for output layers] Compute delta for all neurons in the this (output) layer

Parameters:

targetValues -

updateDelta

public void updateDelta(PropParameter param)

Update delta from neurons in the previous layers

updateWeight

public void updateWeight(PropParameter param)