org.neuroph.contrib.learning

Class SoftMax

java.lang.Object

org.neuroph.core.transfer.TransferFunction

org.neuroph.contrib.learning.SoftMax

All Implemented Interfaces:

java.io.Serializable

public class SoftMax
extends TransferFunction

Activation function which enforces that output neurons have probability distribution (sum of all outputs is one)

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
SoftMax(Layer layer)

Method Summary

Modifier and Type	Method and Description
double	getDerivative(double net) Returns the first derivative of this function.
double	getOutput(double netInput) Returns the ouput of this function.

Methods inherited from class java.lang.Object

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

Constructor Detail

SoftMax

public SoftMax(Layer layer)

Method Detail

getOutput

public double getOutput(double netInput)

Description copied from class: TransferFunction

Returns the ouput of this function.

Specified by:

getOutput in class TransferFunction

Parameters:

netInput - total input

getDerivative

public double getDerivative(double net)

Description copied from class: TransferFunction

Returns the first derivative of this function. Note: should this method should be abstract?

Overrides:

getDerivative in class TransferFunction

Parameters:

net - total input