org.joone.engine.extenders

Class LearnerExtender

java.lang.Object

org.joone.engine.extenders.LearnerExtender

Direct Known Subclasses:

DeltaRuleExtender, GradientExtender, UpdateWeightExtender

public abstract class LearnerExtender
extends java.lang.Object

This abstract class describes the methods that any learner extender must provide.

Constructor Summary

Constructors

Constructor and Description
LearnerExtender()

Method Summary

Modifier and Type	Method and Description
boolean	isEnabled() Checks if the learner extender is enabled.
abstract void	postBiasUpdate(double[] currentGradientOuts) Gives extenders a change to do some post-computing after the biases are updated.
abstract void	postWeightUpdate(double[] currentPattern, double[] currentInps) Gives extenders a change to do some post-computing after the weights are updated.
abstract void	preBiasUpdate(double[] currentGradientOuts) Gives extenders a change to do some pre-computing before the biases are updated.
abstract void	preWeightUpdate(double[] currentPattern, double[] currentInps) Gives extenders a change to do some pre-computing before the weights are updated.
void	setEnabled(boolean aMode) Sets the mode of this extender.
void	setLearner(ExtendableLearner aLearner) Sets the learner.

Methods inherited from class java.lang.Object

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

Constructor Detail

LearnerExtender

public LearnerExtender()

Method Detail

setLearner

public void setLearner(ExtendableLearner aLearner)

Sets the learner. This way the extender has a reference to the learner.

Parameters:

aLearner - the learner this object is extending.

isEnabled

public boolean isEnabled()

Checks if the learner extender is enabled.

Returns:

true if the extender is enabled, false otherwise.

setEnabled

public void setEnabled(boolean aMode)

Sets the mode of this extender.

Parameters:

aMode - true for enabled, false for disabled.

preBiasUpdate

public abstract void preBiasUpdate(double[] currentGradientOuts)

Gives extenders a change to do some pre-computing before the biases are updated.

Parameters:

currentGradientOuts - the back propagated gradients.

postBiasUpdate

public abstract void postBiasUpdate(double[] currentGradientOuts)

Gives extenders a change to do some post-computing after the biases are updated.

Parameters:

currentGradientOuts - the back propagated gradients.

preWeightUpdate

public abstract void preWeightUpdate(double[] currentPattern,
                                     double[] currentInps)

Gives extenders a change to do some pre-computing before the weights are updated.

Parameters:

currentPattern - the back propagated gradients.

currentInps - the forwarded input.

postWeightUpdate

public abstract void postWeightUpdate(double[] currentPattern,
                                      double[] currentInps)

Gives extenders a change to do some post-computing after the weights are updated.

Parameters:

currentPattern - the back propagated gradients.

currentInps - the forwarded input.