jsat.classifiers.neuralnetwork

Class BackPropagationNet.ActivationFunction

java.lang.Object

jsat.classifiers.neuralnetwork.BackPropagationNet.ActivationFunction

All Implemented Interfaces:

Serializable, Function

Enclosing class:

BackPropagationNet

public abstract static class BackPropagationNet.ActivationFunction
extends Object
implements Function

The neural network needs an activation function for the neurons that is used to predict from inputs and train the network by propagating the errors back through the network.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
ActivationFunction()

Method Summary

Modifier and Type	Method and Description
double	f(double... x)
double	f(Vec x)
abstract Function	getD() Returns the function object for the derivative of this activation function.
abstract double	max() The maximum possible response value
abstract double	min() The minimum possible response value
abstract double	response(double x) Computes the response of the response of this activation function on the given input value

Methods inherited from class java.lang.Object

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

Constructor Detail

ActivationFunction

public ActivationFunction()

Method Detail

response

public abstract double response(double x)

Computes the response of the response of this activation function on the given input value

Parameters:

x - the input value

Returns:

the response value

min

public abstract double min()

The minimum possible response value

Returns:

the min value

max

public abstract double max()

The maximum possible response value

Returns:

the max value

getD

public abstract Function getD()

Returns the function object for the derivative of this activation function. The derivative must be calculated using only the output of the response when given the original input. Meaning: given an input x, the value of f'(x) must be computable as g(f(x))

Returns:

the function for computing the derivative of the response

f

public double f(double... x)

Specified by:

f in interface Function

f

public double f(Vec x)

Specified by:

f in interface Function