edu.rice.comp440.net

Class NeuralNetwork

java.lang.Object

edu.rice.comp440.net.NeuralNetwork

All Implemented Interfaces:

java.io.Serializable

public class NeuralNetwork
extends java.lang.Object
implements java.io.Serializable

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
HiddenUnit[][]	hidden
InputUnit[]	input
static java.util.Random	random

Constructor Summary

Constructors

Constructor and Description
NeuralNetwork(int input, int[] hidden) Builds a neural network with the given number of input units, hidden units, and output units.
NeuralNetwork(NeuralNetwork net) Builds a neural network based on the provided network and copies the weights of the provided network into the new one.

Method Summary

Methods

Modifier and Type	Method and Description
double[]	getValue(double[] input) Calculates the network value given the provided input
static NeuralNetwork	readFrom(java.lang.String filename) Method which reads and returns a network from the given file
void	writeTo(java.lang.String filename) Method which writes this network to the given file

Methods inherited from class java.lang.Object

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

Field Detail

random

public static final java.util.Random random

input

public InputUnit[] input

hidden

public HiddenUnit[][] hidden

Constructor Detail

NeuralNetwork

public NeuralNetwork(int input,
             int[] hidden)

Builds a neural network with the given number of input units, hidden units, and output units. Thus, calling new NeuralNetwork(10, new int[] {20, 5}); creates a neural network with 10 input units, a layer of 20 hidden units, and then 5 output units.

Parameters:

input - The number of input units

hidden - The number of hidden units, as well as the number of layers

NeuralNetwork

public NeuralNetwork(NeuralNetwork net)

Builds a neural network based on the provided network and copies the weights of the provided network into the new one.

Parameters:

net - The network to base it off of

Method Detail

getValue

public double[] getValue(double[] input)

Calculates the network value given the provided input

Parameters:

input - The input to check

Returns:

The network value from this input

writeTo

public void writeTo(java.lang.String filename)
             throws java.io.IOException

Method which writes this network to the given file

Parameters:

file - The file to write to

Throws:

java.io.IOException

readFrom

public static NeuralNetwork readFrom(java.lang.String filename)
                              throws java.io.IOException,
                                     java.lang.ClassNotFoundException

Method which reads and returns a network from the given file

Parameters:

filename - The file to read from

Throws:

java.io.IOException

java.lang.ClassNotFoundException