Package libai.nn

Class NeuralNetwork

java.lang.Object

libai.nn.NeuralNetwork

All Implemented Interfaces:

java.io.Serializable

Direct Known Subclasses:

SupervisedLearning, UnsupervisedLearning

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

Neural network abstraction. Provides the methods to train, simulate and calculate the error.

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field	Description
protected Plotter	plotter
protected ProgressDisplay	progress
protected java.util.Random	random

Constructor Summary

Constructors

Constructor	Description
NeuralNetwork()
NeuralNetwork(java.util.Random rand)

Method Summary

Modifier and Type	Method	Description
double	error(Column[] patterns, Column[] answers)	Calculates from a set of patterns.
double	error(Column[] patterns, Column[] answers, int offset, int length)	Calculates the mean quadratic error.
static double	euclideanDistance2(double[] a, double[] b)	Calculates the square Euclidean distance between two vectors.
static double	euclideanDistance2(Column a, Column b)	Calculates the square Euclidean distance between two column matrix.
static double	gaussian(double u2, double sigma)	Calculates the Gaussian function with standard deviation sigma and input parameter u^2
static java.util.Random	getDefaultRandomGenerator()
Plotter	getPlotter()
ProgressDisplay	getProgressBar()
protected void	initializeProgressBar(int maximum)
static <NN extends NeuralNetwork> NN	open(java.io.File file)
static <NN extends NeuralNetwork> NN	open(java.io.InputStream input)
static <NN extends NeuralNetwork> NN	open(java.lang.String path)
boolean	save(java.lang.String path)	Saves the neural network to the file in the given path
void	setPlotter(Plotter plotter)
void	setProgressBar(ProgressDisplay pb)	Sets a ProgressDisplay to the NeuralNetwork.
abstract Column	simulate(Column pattern)	Calculates the output for the pattern.
abstract void	simulate(Column pattern, Column result)	Calculates the output for the pattern and left the result in result.
void	train(Column[] patterns, Column[] answers, double alpha, int epochs)	Alias of train(patterns, answers, alpha, epochs, 0, patterns.length, 1.e-5).
void	train(Column[] patterns, Column[] answers, double alpha, int epochs, int offset, int length)	Alias of train(patterns, answers, alpha, epochs, offset, length, 1.e-5).
abstract void	train(Column[] patterns, Column[] answers, double alpha, int epochs, int offset, int length, double minerror)	Trains this neural network with the list of patterns and the expected answers.

Methods inherited from class java.lang.Object

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

Field Detail

random

protected final java.util.Random random

plotter

protected transient Plotter plotter

progress

protected transient ProgressDisplay progress

Constructor Detail

NeuralNetwork

public NeuralNetwork()

NeuralNetwork

public NeuralNetwork(java.util.Random rand)

Method Detail

getDefaultRandomGenerator

public static final java.util.Random getDefaultRandomGenerator()

open

public static final <NN extends NeuralNetwork> NN open(java.lang.String path)
                                                throws java.io.IOException,
                                                       java.lang.ClassNotFoundException

Throws:

java.io.IOException

java.lang.ClassNotFoundException

open

public static final <NN extends NeuralNetwork> NN open(java.io.File file)
                                                throws java.io.IOException,
                                                       java.lang.ClassNotFoundException

Throws:

java.io.IOException

java.lang.ClassNotFoundException

open

public static final <NN extends NeuralNetwork> NN open(java.io.InputStream input)
                                                throws java.io.IOException,
                                                       java.lang.ClassNotFoundException

Throws:

java.io.IOException

java.lang.ClassNotFoundException

euclideanDistance2

public static double euclideanDistance2(double[] a,
                                        double[] b)

Calculates the square Euclidean distance between two vectors.

NOTE: Assertions of the dimensions are made with assert statement. You must enable this on runtime to be effective.

Parameters:

a - Vector a.

b - Vector b.

Returns:

The square Euclidean distance.

euclideanDistance2

public static double euclideanDistance2(Column a,
                                        Column b)

Calculates the square Euclidean distance between two column matrix.

NOTE: Assertions of the dimensions are made with assert statement. You must enable this on runtime to be effective.

Parameters:

a - Column matrix a.

b - Column matrix b.

Returns:

The square Euclidean distance.

gaussian

public static double gaussian(double u2,
                              double sigma)

Calculates the Gaussian function with standard deviation sigma and input parameter u^2

Parameters:

u2 - u2

sigma - sigma

Returns:

e^(-u^2/2.sigma)

getPlotter

public Plotter getPlotter()

setPlotter

public void setPlotter(Plotter plotter)

getProgressBar

public ProgressDisplay getProgressBar()

setProgressBar

public void setProgressBar(ProgressDisplay pb)

Sets a ProgressDisplay to the NeuralNetwork. The value will go from -epochs to 0, and updated every training epoch.

Note: Classes that implement train(Column[], Column[], double, int, int, int, double) are responsible for this behavior.

Parameters:

pb - ProgressDisplay

train

public abstract void train(Column[] patterns,
                           Column[] answers,
                           double alpha,
                           int epochs,
                           int offset,
                           int length,
                           double minerror)

Trains this neural network with the list of patterns and the expected answers.

Use the learning rate alpha for many epochs. Take length patterns from the position offset until the minerror is reached.

patterns and answers must be arrays of non-null column matrices

Parameters:

patterns - The patterns to be learned.

answers - The expected answers.

alpha - The learning rate.

epochs - The maximum number of iterations

offset - The first pattern position

length - How many patterns will be used.

minerror - The minimal error expected.

train

public void train(Column[] patterns,
                  Column[] answers,
                  double alpha,
                  int epochs)

Alias of train(patterns, answers, alpha, epochs, 0, patterns.length, 1.e-5).

patterns and answers must be arrays of non-null column matrices

Parameters:

patterns - The patterns to be learned.

answers - The expected answers.

alpha - The learning rate.

epochs - The maximum number of iterations

See Also:

train(Column[], Column[], double, int, int, int, double)

train

public void train(Column[] patterns,
                  Column[] answers,
                  double alpha,
                  int epochs,
                  int offset,
                  int length)

Alias of train(patterns, answers, alpha, epochs, offset, length, 1.e-5).

patterns and answers must be arrays of non-null column matrices

Parameters:

patterns - The patterns to be learned.

answers - The expected answers.

alpha - The learning rate.

epochs - The maximum number of iterations

offset - The first pattern position

length - How many patterns will be used.

See Also:

train(Column[], Column[], double, int, int, int, double)

simulate

public abstract Column simulate(Column pattern)

Calculates the output for the pattern.

Parameters:

pattern - Pattern to use as input.

Returns:

The output for the neural network.

simulate

public abstract void simulate(Column pattern,
                              Column result)

Calculates the output for the pattern and left the result in result.

Parameters:

pattern - Pattern to use as input.

result - The output for the input.

save

public boolean save(java.lang.String path)

Saves the neural network to the file in the given path

Parameters:

path - The path for the output file.

Returns:

true if the file can be created and written, false otherwise.

error

public double error(Column[] patterns,
                    Column[] answers)

Calculates from a set of patterns. Alias of error(patterns, answers, 0, patterns.length)

patterns and answers must be arrays of non-null column matrices

Parameters:

patterns - The array with the patterns to test

answers - The array with the expected answers for the patterns.

Returns:

The error calculate for the patterns.

See Also:

error(Column[], Column[], int, int)

error

public double error(Column[] patterns,
                    Column[] answers,
                    int offset,
                    int length)

Calculates the mean quadratic error. It is the standard error metric for neural networks. Just a few networks needs a different type of error metric.

patterns and answers must be arrays of non-null column matrices

NOTE: Assertions of the dimensions are made with assert statement. You must enable this on runtime to be effective.

Parameters:

patterns - The array with the patterns to test

answers - The array with the expected answers for the patterns.

offset - The initial position inside the array.

length - How many patterns must be taken from the offset.

Returns:

The mean quadratic error.

initializeProgressBar

protected void initializeProgressBar(int maximum)