org.ai_heuristic.algo.cluster.som.network

Class DefaultNetwork

java.lang.Object

org.ai_heuristic.algo.cluster.som.network.DefaultNetwork

All Implemented Interfaces:

NetworkModel

public class DefaultNetwork
extends java.lang.Object
implements NetworkModel

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.

Constructor Summary

Constructors

Constructor and Description
DefaultNetwork(java.util.ArrayList neuronWeights, TopologyModel topology, FunctionMetric distanceFunction, FunctionLearn learnFunction, FunctionActivate activationFunction) Create network with specified topology, random weight from defined interval and number of inputs
DefaultNetwork(int weightNumber, double[] maxWeight, TopologyModel topology, FunctionMetric distanceFunction, FunctionLearn learnFunction, FunctionActivate activationFunction) Create network with specified topology, random weight from defined interval and number of inputs
DefaultNetwork(java.lang.String fileName, TopologyModel topology, java.lang.String separator, FunctionMetric distanceFunction, FunctionLearn learnFunction, FunctionActivate activationFunction) Create network with specified topology and parameters get from specified file

Method Summary

Modifier and Type	Method and Description
int	bestNeuronIndex(double[] inputVector) Get the index of the best matching neuron for the input vector.
NeuronModel	getNeuron(int neuronNumber) Return specified by number neuron
int	getNumbersOfNeurons() Return number of neurons
TopologyModel	getTopology() Get topology reference
void	networkToFile(java.lang.String fileName) Save network into file
void	setTopology(TopologyModel topology) Set topology
java.lang.String	toString() Returns a string representation of the Coords object

Methods inherited from class java.lang.Object

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

Constructor Detail

DefaultNetwork

public DefaultNetwork(java.util.ArrayList neuronWeights,
                      TopologyModel topology,
                      FunctionMetric distanceFunction,
                      FunctionLearn learnFunction,
                      FunctionActivate activationFunction)
               throws java.lang.Exception

Create network with specified topology, random weight from defined interval and number of inputs

Parameters:

neuronWeights - array lists of weights for each neuron. Of type double[]. These are added as is and not changed.

topology - Topology

distanceFunction - distance function.

learnFunction - reference to learning function

activationFunction - activation function.

Throws:

java.lang.Exception - any error.

DefaultNetwork

public DefaultNetwork(int weightNumber,
                      double[] maxWeight,
                      TopologyModel topology,
                      FunctionMetric distanceFunction,
                      FunctionLearn learnFunction,
                      FunctionActivate activationFunction)
               throws java.lang.Exception

Create network with specified topology, random weight from defined interval and number of inputs

Parameters:

weightNumber - number of weights (inputs)

maxWeight - array with specified max weight values. These are added as a random value up to the allowed maximum.

topology - Topology

distanceFunction - distance function.

learnFunction - reference to learning function

activationFunction - activation function.

Throws:

java.lang.Exception - any error.

DefaultNetwork

public DefaultNetwork(java.lang.String fileName,
                      TopologyModel topology,
                      java.lang.String separator,
                      FunctionMetric distanceFunction,
                      FunctionLearn learnFunction,
                      FunctionActivate activationFunction)
               throws java.lang.Exception

Create network with specified topology and parameters get from specified file

Parameters:

fileName - File Name

topology - Topology

separator - the data value separator.

distanceFunction - distance function.

learnFunction - reference to learning function

activationFunction - activation function.

Throws:

java.lang.Exception - any error.

Method Detail

bestNeuronIndex

public int bestNeuronIndex(double[] inputVector)
                    throws java.lang.Exception

Get the index of the best matching neuron for the input vector.

Specified by:

bestNeuronIndex in interface NetworkModel

Parameters:

inputVector - an input dataset vector.

Returns:

the neuron with the best resulting value.

Throws:

java.lang.Exception - any error.

getNeuron

public NeuronModel getNeuron(int neuronNumber)

Return specified by number neuron

Specified by:

getNeuron in interface NetworkModel

Parameters:

neuronNumber - neuron number

Returns:

Neuorn

getNumbersOfNeurons

public int getNumbersOfNeurons()

Return number of neurons

Specified by:

getNumbersOfNeurons in interface NetworkModel

Returns:

number of neurons

getTopology

public TopologyModel getTopology()

Get topology reference

Specified by:

getTopology in interface NetworkModel

Returns:

Topology

See Also:

TopologyModel

setTopology

public void setTopology(TopologyModel topology)

Set topology

Specified by:

setTopology in interface NetworkModel

Parameters:

topology - Topology

toString

public java.lang.String toString()

Returns a string representation of the Coords object

Overrides:

toString in class java.lang.Object

Returns:

Returns a string representation of the Coords object

networkToFile

public void networkToFile(java.lang.String fileName)

Save network into file

Parameters:

fileName - File Name