com.imsl.datamining.neural
Class OutputLayer

java.lang.Object
  com.imsl.datamining.neural.Layer
      com.imsl.datamining.neural.OutputLayer

All Implemented Interfaces:

Serializable

public class OutputLayer

extends Layer

Output layer in a neural network. An empty OutputLayer is automatically created by FeedForwardNetwork.

See Also:

Network, Serialized Form

Field Summary

static long

serialVersionUID

Method Summary

Perceptron	createPerceptron() Creates a Perceptron in this Layer of the neural network.
Perceptron	createPerceptron(Activation activation, double bias) Creates a Perceptron in this Layer with a specified Activation and bias.
Perceptron[]	createPerceptrons(int n) Creates a number of Perceptrons in this Layer of the neural network.
Perceptron[]	createPerceptrons(int n, Activation activation, double bias) Creates a number of Perceptrons in this Layer with specified activation and bias.
Node[]	getNodes() Return the Perceptrons in the OutputLayer.

Methods inherited from class com.imsl.datamining.neural.Layer

addNode, getIndex

Methods inherited from class java.lang.Object

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

Field Detail

serialVersionUID

public static final long serialVersionUID

See Also:

Constant Field Values

Method Detail

createPerceptron

public Perceptron createPerceptron()

Creates a Perceptron in this Layer of the neural network. By default, the created Perceptron uses the linear activation function and has an initial bias value of zero.

createPerceptron

public Perceptron createPerceptron(Activation activation,
                                   double bias)

Creates a Perceptron in this Layer with a specified Activation and bias.

Parameters:

activation - The Activation object which specifies the action function to be used.

bias - A double which specifies the initial value for the bias for this Perceptron.

createPerceptrons

public Perceptron[] createPerceptrons(int n)

Creates a number of Perceptrons in this Layer of the neural network. By default, they will use linear activation and a zero initial bias.

Parameters:

n - An int which specifies the number of Perceptrons to be created in this layer.

Returns:

An array containing the created Perceptrons.

createPerceptrons

public Perceptron[] createPerceptrons(int n,
                                      Activation activation,
                                      double bias)

Creates a number of Perceptrons in this Layer with specified activation and bias.

Parameters:

n - An int which specifies the number of Perceptrons to be created.

activation - The Activation object which indicates the action function to be used.

bias - A double which specifies the initial bias for the Perceptrons.

Returns:

An array containing the created Perceptrons.

getNodes

public Node[] getNodes()

Return the Perceptrons in the OutputLayer.

This method overides the method in Layer to return the Perceptrons in an OutputPerceptron array.

Overrides:

getNodes in class Layer

Returns:

An OutputPerceptron[] array containing the Nodes in the OutputLayer.