neural.net
Interface ActivationFunction

All Superinterfaces:

java.io.Serializable

All Known Implementing Classes:

ActivationFunctionSigmoid

public interface ActivationFunction
extends java.io.Serializable

Interface to implement for any Activation Function to be used with a Neural Network

Version:

1.0.0

Author:

Jonathan Reimels

Method Summary

double	activate(double input) The activation function to execute within a neuron
double	derivative(double input) The derivative of the Activation Function to use in Back-Propogation

Method Detail

activate

double activate(double input)

The activation function to execute within a neuron

Parameters:

input - - the input to process, should be the sum of weights and input to a neuron

Returns:

output

derivative

double derivative(double input)

The derivative of the Activation Function to use in Back-Propogation

Parameters:

input - - the input to process through the derivative function

Returns:

output