org.encog.neural.error

Class CrossEntropyErrorFunction

java.lang.Object

org.encog.neural.error.CrossEntropyErrorFunction

All Implemented Interfaces:

ErrorFunction

public class CrossEntropyErrorFunction
extends java.lang.Object
implements ErrorFunction

Implements a cross entropy error function. This can be used with backpropagation to sometimes provide better performance than the standard linear error function.

Constructor Summary

Constructors

Constructor and Description
CrossEntropyErrorFunction()

Method Summary

Modifier and Type	Method and Description
void	calculateError(ActivationFunction af, double[] b, double[] a, double[] ideal, double[] actual, double[] error, double derivShift, double significance) Calculate the error.

Methods inherited from class java.lang.Object

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

Constructor Detail

CrossEntropyErrorFunction

public CrossEntropyErrorFunction()

Method Detail

calculateError

public void calculateError(ActivationFunction af,
                           double[] b,
                           double[] a,
                           double[] ideal,
                           double[] actual,
                           double[] error,
                           double derivShift,
                           double significance)

Calculate the error.

Specified by:

calculateError in interface ErrorFunction

Parameters:

af - The activation function used at the output layer.

b - The number to calculate the derivative of, the number "before" the activation function was applied.

a - The number "after" an activation function has been applied.

ideal - The ideal values.

actual - The actual values.

error - The resulting error values.

derivShift - The amount to shift af derivativeFunction by

significance - Weighting to apply to ideal[i] - actual[i]