org.encog.neural.networks.training.strategy
Class ResetStrategy

java.lang.Object
  org.encog.neural.networks.training.strategy.ResetStrategy

All Implemented Interfaces:

Strategy

public class ResetStrategy
extends java.lang.Object
implements Strategy

The reset strategy will reset the weights if the neural network fails to fall below a specified error by a specified number of cycles. This can be useful to throw out initially "bad/hard" random initializations of the weight matrix.

Author:

jheaton

Constructor Summary

ResetStrategy(double required, int cycles)
Construct a reset strategy.

Method Summary

void	init(Train train) Initialize this strategy.
void	postIteration() Called just after a training iteration.
void	preIteration() Called just before a training iteration.

Methods inherited from class java.lang.Object

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

Constructor Detail

ResetStrategy

public ResetStrategy(double required,
                     int cycles)

Construct a reset strategy. The error rate must fall below the required rate in the specified number of cycles, or the neural network will be reset to random weights and thresholds.

Parameters:

required - The required error rate.

cycles - The number of cycles to reach that rate.

Method Detail

init

public void init(Train train)

Initialize this strategy.

Specified by:

init in interface Strategy

Parameters:

train - The training algorithm.

postIteration

public void postIteration()

Called just after a training iteration.

Specified by:

postIteration in interface Strategy

preIteration

public void preIteration()

Called just before a training iteration.

Specified by:

preIteration in interface Strategy