org.ai_heuristic.algo.cluster.som.kohonen

Class WTMLearningFunction

java.lang.Object

org.ai_heuristic.algo.cluster.som.kohonen.WTMLearningFunction

All Implemented Interfaces:

LearningFunction

Direct Known Subclasses:

WTMLearningFunctionWithTired

public class WTMLearningFunction
extends java.lang.Object
implements LearningFunction

Updated version of the source code for the kohonen self-organising map at http://jknnl.sourceforge.net/ Some classes or packages may have been moved or changed. See above for the copyright notices.

Author:

Kieran Greer.

Field Summary

Fields

Modifier and Type	Field and Description
protected FunctionLearn	functionalModel Reference to function model
protected LearningDataModel	learningData Reference to learning data
protected int	maxIteration Max number of iteration
protected FunctionMetric	metrics Reference to metrics
protected NeighbourhoodFunctionModel	neighbourhoodFunction Reference to neighbourhood function model
protected NetworkModel	networkModel Reference to network model
protected TopologyModel	topology Reference to topology model

Constructor Summary

Constructors

Constructor and Description
WTMLearningFunction(NetworkModel networkModel, int maxIteration, FunctionMetric metrics, LearningDataModel learningData, FunctionLearn functionalModel, NeighbourhoodFunctionModel neighbourhoodFunction) Creates a new instance of WTMLearningFunction

Method Summary

Modifier and Type	Method and Description
protected void	changeNeuronWeight(int neuronNumber, double[] vector, double factor, double influence) Change neuron weights for specified neuron number, iteration, input data vector and distance and distance to winning neuron
void	changeWeight(int neuronNumber, double[] vector, int iteration) Change specified neuron weight
protected int	getBestNeuron(double[] vector) Return number of best neuron for specified input vector
FunctionLearn	getFunctionalModel() Return function model
LearningDataModel	getLearningData() Return reference to learning data
int	getMaxIteration() Return maximal number of iteration
FunctionMetric	getMetrics() Return metrics
NeighbourhoodFunctionModel	getNeighbourhoodFunction() Return reference to neighbourhood function
NetworkModel	getNetworkModel() Return network model
void	learn() Start learning process
void	setFunctionalModel(FunctionLearn functionalModel) Set functional learning factor model
void	setLearningData(LearningDataModel learningData) Set reference to learning data
void	setMaxIteration(int maxIteration) Set max iteration
void	setMetrics(FunctionMetric metrics) Set metrics
void	setNeighboorhoodFunction(NeighbourhoodFunctionModel neighbourhoodFunction) Set reference to neighbourhood function
void	setNetworkModel(NetworkModel networkModel) Set network model

Methods inherited from class java.lang.Object

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

Field Detail

metrics

protected FunctionMetric metrics

Reference to metrics

networkModel

protected NetworkModel networkModel

Reference to network model

maxIteration

protected int maxIteration

Max number of iteration

learningData

protected LearningDataModel learningData

Reference to learning data

functionalModel

protected FunctionLearn functionalModel

Reference to function model

topology

protected TopologyModel topology

Reference to topology model

neighbourhoodFunction

protected NeighbourhoodFunctionModel neighbourhoodFunction

Reference to neighbourhood function model

Constructor Detail

WTMLearningFunction

public WTMLearningFunction(NetworkModel networkModel,
                           int maxIteration,
                           FunctionMetric metrics,
                           LearningDataModel learningData,
                           FunctionLearn functionalModel,
                           NeighbourhoodFunctionModel neighbourhoodFunction)

Creates a new instance of WTMLearningFunction

Parameters:

networkModel - reference to network Model

maxIteration - max number of iteration

metrics - reference to metrics

learningData - reference to learning data

functionalModel - reference to functional Model

neighbourhoodFunction - reference to neighbourhood function. If null this is set to a FractionNeighbourhoodFunction.

Method Detail

setNeighboorhoodFunction

public void setNeighboorhoodFunction(NeighbourhoodFunctionModel neighbourhoodFunction)

Set reference to neighbourhood function

Parameters:

neighbourhoodFunction - reference to neighbourhood function

getNeighbourhoodFunction

public NeighbourhoodFunctionModel getNeighbourhoodFunction()

Return reference to neighbourhood function

Returns:

reference to neighbourhood function

getMetrics

public FunctionMetric getMetrics()

Return metrics

Returns:

metrics

See Also:

FunctionMetric

setMetrics

public void setMetrics(FunctionMetric metrics)

Set metrics

Parameters:

metrics - metrics

setNetworkModel

public void setNetworkModel(NetworkModel networkModel)

Set network model

Parameters:

networkModel - network model

getNetworkModel

public NetworkModel getNetworkModel()

Return network model

Returns:

network model

setMaxIteration

public void setMaxIteration(int maxIteration)

Set max iteration

Parameters:

maxIteration - max iteration

getMaxIteration

public int getMaxIteration()

Return maximal number of iteration

Returns:

maximal number of iteration

setLearningData

public void setLearningData(LearningDataModel learningData)

Set reference to learning data

Parameters:

learningData - reference to learning data

getLearningData

public LearningDataModel getLearningData()

Return reference to learning data

Returns:

reference to learning data

setFunctionalModel

public void setFunctionalModel(FunctionLearn functionalModel)

Set functional learning factor model

Parameters:

functionalModel - functional learning factor model

getFunctionalModel

public FunctionLearn getFunctionalModel()

Return function model

Returns:

function model

getBestNeuron

protected int getBestNeuron(double[] vector)
                     throws java.lang.Exception

Return number of best neuron for specified input vector

Parameters:

vector - input vector

Returns:

NeuronModelnumber

Throws:

java.lang.Exception - any error.

changeWeight

public void changeWeight(int neuronNumber,
                         double[] vector,
                         int iteration)
                  throws java.lang.Exception

Change specified neuron weight

Parameters:

neuronNumber - neuron Number

vector - input vector

iteration - iteration number

Throws:

java.lang.Exception - any error.

changeNeuronWeight

protected void changeNeuronWeight(int neuronNumber,
                                  double[] vector,
                                  double factor,
                                  double influence)
                           throws java.lang.Exception

Change neuron weights for specified neuron number, iteration, input data vector and distance and distance to winning neuron

Parameters:

neuronNumber - neuron number

vector - input vector

factor - the iteration number or factor.

influence - cells in neighbourhood have a reduced influence.

Throws:

java.lang.Exception - any error.

learn

public void learn()
           throws java.lang.Exception

Start learning process

Specified by:

learn in interface LearningFunction

Throws:

java.lang.Exception - any error.