keel.Algorithms.Neural_Networks.NNEP_Common

Class AbstractNeuralNetSpecies<I extends NeuralNetIndividual>

java.lang.Object

keel.Algorithms.Neural_Networks.NNEP_Common.AbstractNeuralNetSpecies<I>

Type Parameters:

I - Type of represented individuals

All Implemented Interfaces:

INeuralNetSpecies<I>

Direct Known Subclasses:

NeuralNetIndividualSpecies

public abstract class AbstractNeuralNetSpecies<I extends NeuralNetIndividual>
extends java.lang.Object
implements INeuralNetSpecies<I>

Abstract implementation for INeuralNetSpecies.

Since:

JDK1.5

Version:

0.1

Author:

Written by Pedro Antonio Gutierrez Penia (University of Cordoba) 16/7/2007

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean[]	biased Boolean indicating if each linked layer of neural nets are biased
protected int[]	initialMaxNofneurons Initial number of neurons of each LinkedLayer of the neural nets
protected java.lang.String[]	initiator Initiator of each LinkedLayer of the neural nets
protected java.lang.String[][]	initiatorNeuronTypes Initiator of neurons of each HibridLayer of the neural nets
protected int[]	maxNofneurons Maximum number of neurons of each LinkedLayer of the neural nets
protected int[]	minNofneurons Minimum number of neurons of each LinkedLayer of the neural nets
protected java.lang.String	neuralNetType Type of neuralnets
protected java.lang.String[][]	neuronTypes Types of each neuron for hibrid layers
protected int	nOfHiddenLayers Number of hidden layers of the neural nets
protected int	nOfInputs Number of inputs of the neural nets
protected int	nOfOutputs Number of outputs
protected double[][]	percentages Percentages of each neuron type for hibrid layers
protected java.lang.String[]	type Type of each LinkedLayer of the neural nets
protected Interval[][]	weightRanges Weight ranges of each LinkedLayer of the neural nets

Constructor Summary

Constructors

Constructor and Description
AbstractNeuralNetSpecies() Empty constructor

Method Summary

Modifier and Type	Method and Description
int	getHiddenLayerInitialMaxNofneurons(int index) Returns initial maximum number of neurons of a hidden layer
java.lang.String	getHiddenLayerInitiator(int index) Returns initiator of neurons of a hidden layer
int	getHiddenLayerMaxNofneurons(int index) Returns maximum number of neurons of a hidden layer
int	getHiddenLayerMinNofneurons(int index) Returns minimum number of neurons of a hidden layer
java.lang.String	getHiddenLayerType(int index) Returns type of neurons of a hidden layer
Interval	getHiddenLayerWeightRange(int index, int indexRange) Returns weight range of a hidden layer
java.lang.String[]	getInitiatorNeuronTypes(int index) Returns an array of initiators of neurons of hibrid layers
java.lang.String	getNeuralNetType() Returns a neural net type
java.lang.String[]	getNeuronTypes(int index) Returns an array of neuron types of a concrete layer (this is an hibrid layer)
int	getNOfHiddenLayers() Returns number of hidden layers of the neural nets
int	getNOfInputs() Returns number of inputs of the neural nets
int	getNOfOutputs() Returns number of outputs of the neural nets
java.lang.String	getOutputLayerInitiator() Returns initiator of neurons of the output layer
java.lang.String	getOutputLayerType() Returns type of neurons of the output layer
Interval	getOutputLayerWeightRange(int indexRange) Returns weight range of the output layer
double[]	getPercentages(int index) Returns an array of percentages of a concrete layer (this is an hibrid layer)
boolean	isHiddenLayerBiased(int index) Returns a boolean indicating if a hidden layer is biased
boolean	isOutputLayerBiased() Returns a boolean indicating if output layer is biased

Methods inherited from class java.lang.Object

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

Methods inherited from interface keel.Algorithms.Neural_Networks.NNEP_Common.INeuralNetSpecies

createGenotype, createIndividual

Field Detail

neuralNetType

protected java.lang.String neuralNetType

Type of neuralnets

nOfInputs

protected int nOfInputs

Number of inputs of the neural nets

nOfHiddenLayers

protected int nOfHiddenLayers

Number of hidden layers of the neural nets

nOfOutputs

protected int nOfOutputs

Number of outputs

weightRanges

protected Interval[][] weightRanges

Weight ranges of each LinkedLayer of the neural nets

maxNofneurons

protected int[] maxNofneurons

Maximum number of neurons of each LinkedLayer of the neural nets

minNofneurons

protected int[] minNofneurons

Minimum number of neurons of each LinkedLayer of the neural nets

initialMaxNofneurons

protected int[] initialMaxNofneurons

Initial number of neurons of each LinkedLayer of the neural nets

type

protected java.lang.String[] type

Type of each LinkedLayer of the neural nets

initiator

protected java.lang.String[] initiator

Initiator of each LinkedLayer of the neural nets

biased

protected boolean[] biased

Boolean indicating if each linked layer of neural nets are biased

neuronTypes

protected java.lang.String[][] neuronTypes

Types of each neuron for hibrid layers

percentages

protected double[][] percentages

Percentages of each neuron type for hibrid layers

initiatorNeuronTypes

protected java.lang.String[][] initiatorNeuronTypes

Initiator of neurons of each HibridLayer of the neural nets

Constructor Detail

AbstractNeuralNetSpecies

public AbstractNeuralNetSpecies()

Empty constructor

Method Detail

getNeuralNetType

public java.lang.String getNeuralNetType()

Returns a neural net type

Returns:

int Number of hidden layers

getNOfHiddenLayers

public int getNOfHiddenLayers()

Returns number of hidden layers of the neural nets

Specified by:

getNOfHiddenLayers in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Returns:

int Number of hidden layers

getNOfInputs

public int getNOfInputs()

Returns number of inputs of the neural nets

Specified by:

getNOfInputs in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Returns:

int Number of inputs

getNOfOutputs

public int getNOfOutputs()

Returns number of outputs of the neural nets

Specified by:

getNOfOutputs in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Returns:

int Number of outputs

getHiddenLayerWeightRange

public Interval getHiddenLayerWeightRange(int index,
                                          int indexRange)

Returns weight range of a hidden layer

Specified by:

getHiddenLayerWeightRange in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

index - Index of the desired hidden layer

indexRange - Index of the desired range into the layer (useful for hibrid layer)

Returns:

Interval Weight range

getOutputLayerWeightRange

public Interval getOutputLayerWeightRange(int indexRange)

Returns weight range of the output layer

Specified by:

getOutputLayerWeightRange in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

indexRange - Index of the desired range into the layer (useful for hibrid layer)

Returns:

Interval Weight range

getHiddenLayerMinNofneurons

public int getHiddenLayerMinNofneurons(int index)

Returns minimum number of neurons of a hidden layer

Specified by:

getHiddenLayerMinNofneurons in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

index - Index of the desired hidden layer

Returns:

int Minimum number of neurons

getHiddenLayerMaxNofneurons

public int getHiddenLayerMaxNofneurons(int index)

Returns maximum number of neurons of a hidden layer

Specified by:

getHiddenLayerMaxNofneurons in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

index - Index of the desired hidden layer

Returns:

int Maximum number of neurons

getHiddenLayerInitialMaxNofneurons

public int getHiddenLayerInitialMaxNofneurons(int index)

Returns initial maximum number of neurons of a hidden layer

Specified by:

getHiddenLayerInitialMaxNofneurons in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

index - Index of the desired hidden layer

Returns:

int Initial maximum number of neurons

getHiddenLayerType

public java.lang.String getHiddenLayerType(int index)

Returns type of neurons of a hidden layer

Specified by:

getHiddenLayerType in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

index - Index of the desired hidden layer

Returns:

String Type of neurons

getOutputLayerType

public java.lang.String getOutputLayerType()

Returns type of neurons of the output layer

Specified by:

getOutputLayerType in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Returns:

String Type of neurons

getHiddenLayerInitiator

public java.lang.String getHiddenLayerInitiator(int index)

Returns initiator of neurons of a hidden layer

Specified by:

getHiddenLayerInitiator in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

index - Index of the desired hidden layer

Returns:

String Initiator of neurons

getOutputLayerInitiator

public java.lang.String getOutputLayerInitiator()

Returns initiator of neurons of the output layer

Specified by:

getOutputLayerInitiator in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Returns:

String Initiator of neurons

isHiddenLayerBiased

public boolean isHiddenLayerBiased(int index)

Returns a boolean indicating if a hidden layer is biased

Specified by:

isHiddenLayerBiased in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

index - Index of the desired hidden layer

Returns:

boolean Is hidden layer biased?

isOutputLayerBiased

public boolean isOutputLayerBiased()

Returns a boolean indicating if output layer is biased

Specified by:

isOutputLayerBiased in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Returns:

boolean Is output layer biased?

getNeuronTypes

public java.lang.String[] getNeuronTypes(int index)

Returns an array of neuron types of a concrete layer (this is an hibrid layer)

Specified by:

getNeuronTypes in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

index - Index of the desired hidden layer

Returns:

String[] Array of neurons types

getPercentages

public double[] getPercentages(int index)

Returns an array of percentages of a concrete layer (this is an hibrid layer)

Specified by:

getPercentages in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

index - Index of the desired hidden layer

Returns:

double[] Array of percentages

getInitiatorNeuronTypes

public java.lang.String[] getInitiatorNeuronTypes(int index)

Returns an array of initiators of neurons of hibrid layers

Specified by:

getInitiatorNeuronTypes in interface INeuralNetSpecies<I extends NeuralNetIndividual>

Parameters:

index - Index of the desired hidden layer

Returns:

String[] Array of percentages