nuroko.core

Interface IParameterised

All Superinterfaces:

Cloneable, IModule

All Known Subinterfaces:

IComponent, ITrainable

All Known Implementing Classes:

AComponent, ACompoundComponent, ALayerStack, AOperationComponent, AWeightLayer, Bias, CompoundLayerStack, Dropout, FullWeightLayer, GradientAmplifier, Identity, Join, NeuralNet, Normaliser, Operator, SparseWeightLayer, Sparsifier, Stack, ThinkingOp, TrainingOp

public interface IParameterised
extends IModule

Interface representing a parameterised model. The model is assumed to contain two vectors: - A parameter vector - An accumulated gradient vector Both vectors are mutable. They can be cleared, randomized etc. as desired. Altering the parameter vectors is *not thread safe*. It will not throw an exception, however results may be unpredictable if the model is being concurrently used by another thread.

Author:

Mike

Method Summary

Modifier and Type	Method and Description
IParameterised	clone() Creates a clone of a module, including a deep copy of any mutable state.
mikera.vectorz.AVector	getGradient() Return an AVector referencing the accumulated gradient in this model
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.

Methods inherited from interface nuroko.core.IModule

getModules

Method Detail

getParameterLength

int getParameterLength()

Returns the length of the parameter vector for this model

Returns:

getParameters

mikera.vectorz.AVector getParameters()

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

Returns:

getGradient

mikera.vectorz.AVector getGradient()

Return an AVector referencing the accumulated gradient in this model

Returns:

clone

IParameterised 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

Returns: