org.encog.ml.train.strategy
Class RequiredImprovementStrategy

java.lang.Object
  extended by org.encog.ml.train.strategy.RequiredImprovementStrategy
All Implemented Interfaces:
Strategy

public class RequiredImprovementStrategy
extends Object
implements Strategy

The reset strategy will reset the weights if the neural network fails to improve by the specified amount over a number of cycles.

Author:
jheaton

Constructor Summary
RequiredImprovementStrategy(double required, double threshold, int cycles)
          Construct a reset strategy.
RequiredImprovementStrategy(double required, int cycles)
          Construct a reset strategy.
RequiredImprovementStrategy(int cycles)
          Reset if there is not at least a 1% improvement for 5 cycles.
 
Method Summary
 void init(MLTrain 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

RequiredImprovementStrategy

public RequiredImprovementStrategy(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 bias values.

Parameters:
required - The required error rate.
cycles - The number of cycles to reach that rate.

RequiredImprovementStrategy

public RequiredImprovementStrategy(double required,
                                   double threshold,
                                   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 bias values.

Parameters:
required - The required error rate.
threshold - The accepted threshold, don't reset if error is below this.
cycles - The number of cycles to reach that rate.

RequiredImprovementStrategy

public RequiredImprovementStrategy(int cycles)
Reset if there is not at least a 1% improvement for 5 cycles. Don't reset if below 10%.

Parameters:
cycles -
Method Detail

init

public void init(MLTrain 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


Copyright © 2014. All Rights Reserved.