nuroko.module

Class SparseWeightLayer

java.lang.Object

nuroko.module.ALayer

nuroko.module.AWeightLayer

nuroko.module.SparseWeightLayer

All Implemented Interfaces:

Cloneable, IInput, IInputOutput, IModule, IOutput, IParameterised, IThinker

public final class SparseWeightLayer
extends AWeightLayer

Fully connected weight layer

Author:

Mike

Field Summary

Fields inherited from class nuroko.module.AWeightLayer

BIAS_INITIAL_SCALE, INITIAL_WEIGHT_SCALE, inputLength, outputLength

Constructor Summary

Constructors

Constructor and Description
SparseWeightLayer(int inputLength, int outputLength, int maxLinks)
SparseWeightLayer(SparseWeightLayer wl)

Method Summary

Methods

Modifier and Type	Method and Description
mikera.matrixx.AMatrix	asMatrix()
SparseWeightLayer	clone() Creates a clone of a module, including a deep copy of any mutable state.
List<IModule>	getComponents() Returns a list of sub-components of this module
mikera.vectorz.AVector	getGradient() Return an AVector referencing the accumulated gradient in this model
SparseWeightLayer	getInverse()
int	getLinkCount(int outputIndex)
int	getLinkSource(int outputIndex, int number)
double	getLinkWeight(int outputIndex, int number)
int	getParameterLength() Returns the length of the parameter vector for this model
mikera.vectorz.AVector	getParameters() Return an AVector referring to the parameters in the model.
mikera.indexz.Index	getSourceIndex(int outputIndex)
mikera.vectorz.AVector	getSourceWeights(int outputIndex)
void	initRandom()
void	think(mikera.vectorz.AVector input, mikera.vectorz.AVector output)
void	trainGradient(mikera.vectorz.AVector input, mikera.vectorz.AVector outputGradient, mikera.vectorz.AVector inputGradient, double factor)

Methods inherited from class nuroko.module.AWeightLayer

getInputLength, getOutputLength

Methods inherited from class java.lang.Object

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

Constructor Detail

SparseWeightLayer

public SparseWeightLayer(int inputLength,
                 int outputLength,
                 int maxLinks)

SparseWeightLayer

public SparseWeightLayer(SparseWeightLayer wl)

Method Detail

getComponents

public List<IModule> getComponents()

Description copied from interface: IModule

Returns a list of sub-components of this module

Specified by:

getComponents in interface IModule

Overrides:

getComponents in class AWeightLayer

Returns:

asMatrix

public mikera.matrixx.AMatrix asMatrix()

Specified by:

asMatrix in class AWeightLayer

getParameters

public mikera.vectorz.AVector getParameters()

Description copied from interface: IParameterised

Return an AVector referring to the parameters in the model. Expected to be mutable.

Specified by:

getParameters in interface IParameterised

Specified by:

getParameters in class ALayer

Returns:

getParameterLength

public int getParameterLength()

Description copied from interface: IParameterised

Returns the length of the parameter vector for this model

Specified by:

getParameterLength in interface IParameterised

Specified by:

getParameterLength in class ALayer

Returns:

think

public void think(mikera.vectorz.AVector input,
         mikera.vectorz.AVector output)

Specified by:

think in interface IThinker

Specified by:

think in class ALayer

getGradient

public mikera.vectorz.AVector getGradient()

Description copied from interface: IParameterised

Return an AVector referencing the accumulated gradient in this model

Returns:

trainGradient

public void trainGradient(mikera.vectorz.AVector input,
                 mikera.vectorz.AVector outputGradient,
                 mikera.vectorz.AVector inputGradient,
                 double factor)

Specified by:

trainGradient in class AWeightLayer

clone

public SparseWeightLayer clone()

Description copied from interface: IModule

Creates a clone of a module, including a deep copy of any mutable state.

Specified by:

clone in interface IModule

Specified by:

clone in interface IParameterised

Specified by:

clone in interface IThinker

Specified by:

clone in class AWeightLayer

Returns:

getLinkCount

public int getLinkCount(int outputIndex)

Specified by:

getLinkCount in class AWeightLayer

getLinkWeight

public double getLinkWeight(int outputIndex,
                   int number)

Specified by:

getLinkWeight in class AWeightLayer

getLinkSource

public int getLinkSource(int outputIndex,
                int number)

Specified by:

getLinkSource in class AWeightLayer

initRandom

public void initRandom()

Specified by:

initRandom in class AWeightLayer

getSourceIndex

public mikera.indexz.Index getSourceIndex(int outputIndex)

Specified by:

getSourceIndex in class AWeightLayer

getSourceWeights

public mikera.vectorz.AVector getSourceWeights(int outputIndex)

Specified by:

getSourceWeights in class AWeightLayer

getInverse

public SparseWeightLayer getInverse()

Specified by:

getInverse in class AWeightLayer