com.rapidminer.operator.learner.functions.neuralnet
Class SigmoidFunction

java.lang.Object
  com.rapidminer.operator.learner.functions.neuralnet.ActivationFunction
      com.rapidminer.operator.learner.functions.neuralnet.SigmoidFunction

public class SigmoidFunction
extends ActivationFunction

This function represents a sigmoid activation function by calculating 1 / (1 + exp(- weighted sum). The sigmoid function is usually used for the input and hidden layers and for the output layer for classification problems.

Author:

Ingo Mierswa

Constructor Summary

SigmoidFunction()

Method Summary

double	calculateError(InnerNode node, Example example) Calculates the error.
double	calculateValue(InnerNode node, Example example) Calculates the value.
java.lang.String	getTypeName() Delivers the name of this activation function.

Methods inherited from class com.rapidminer.operator.learner.functions.neuralnet.ActivationFunction

update

Methods inherited from class java.lang.Object

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

Constructor Detail

SigmoidFunction

public SigmoidFunction()

Method Detail

getTypeName

public java.lang.String getTypeName()

Description copied from class: ActivationFunction

Delivers the name of this activation function.

Specified by:

getTypeName in class ActivationFunction

calculateValue

public double calculateValue(InnerNode node,
                             Example example)

Description copied from class: ActivationFunction

Calculates the value.

Specified by:

calculateValue in class ActivationFunction

calculateError

public double calculateError(InnerNode node,
                             Example example)

Description copied from class: ActivationFunction

Calculates the error.

Specified by:

calculateError in class ActivationFunction