Package libai.nn.supervised

Class SVM

java.lang.Object

libai.nn.NeuralNetwork

libai.nn.supervised.SupervisedLearning

libai.nn.supervised.SVM

All Implemented Interfaces:

java.io.Serializable

public class SVM
extends SupervisedLearning

Implementation of the SVM using the SMO algorithm. Based on the original implementation of:
X. Jiang and H. Yu. SVM-JAVA: A Java implementation of the SMO (Sequential Minimal Optimization) for training SVM.
Department of Computer Science and Engineering, Pohang University of Science and Technology (POSTECH), http://iis.hwanjoyu.org/svm-java, 2008. The code was adapted to the data structures and architecture of the libai. Some little optimization was made.

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field	Description
static int	PARAM_C
static int	PARAM_EPSILON
static int	PARAM_TOLERANCE
protected static SymmetricSign	ssign

Fields inherited from class libai.nn.NeuralNetwork

plotter, progress, random

Constructor Summary

Constructors

Constructor	Description
SVM(Kernel _kernel)
SVM(Kernel _kernel, java.util.Random rand)

Method Summary

Modifier and Type	Method	Description
double	error(Column[] patterns, Column[] answers, int offset, int length)	Calculates the mean quadratic error.
void	setTrainingParam(int param, double paramValue)
Column	simulate(Column pattern)	Calculates the output for the pattern.
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, int offset, int length, double minerror)	Trains this neural network with the list of patterns and the expected answers.
protected void	validatePreconditions(Column[] patterns, Column[] answers, int epochs, int offset, int length, double minerror)

Methods inherited from class libai.nn.NeuralNetwork

error, euclideanDistance2, euclideanDistance2, gaussian, getDefaultRandomGenerator, getPlotter, getProgressBar, initializeProgressBar, open, open, open, save, setPlotter, setProgressBar, train, train

Methods inherited from class java.lang.Object

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

Field Detail

PARAM_C

public static final int PARAM_C

See Also:

Constant Field Values

PARAM_EPSILON

public static final int PARAM_EPSILON

See Also:

Constant Field Values

PARAM_TOLERANCE

public static final int PARAM_TOLERANCE

See Also:

Constant Field Values

ssign

protected static final SymmetricSign ssign

Constructor Detail

SVM

public SVM(Kernel _kernel)

SVM

public SVM(Kernel _kernel,
           java.util.Random rand)

Method Detail

setTrainingParam

public void setTrainingParam(int param,
                             double paramValue)

validatePreconditions

protected void validatePreconditions(Column[] patterns,
                                     Column[] answers,
                                     int epochs,
                                     int offset,
                                     int length,
                                     double minerror)

Overrides:

validatePreconditions in class SupervisedLearning

train

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

Description copied from class: NeuralNetwork

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

Specified by:

train in class NeuralNetwork

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.

simulate

public Column simulate(Column pattern)

Description copied from class: NeuralNetwork

Calculates the output for the pattern.

Specified by:

simulate in class NeuralNetwork

Parameters:

pattern - Pattern to use as input.

Returns:

The output for the neural network.

simulate

public void simulate(Column pattern,
                     Column result)

Description copied from class: NeuralNetwork

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

Specified by:

simulate in class NeuralNetwork

Parameters:

pattern - Pattern to use as input.

result - The output for the input.

error

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

Description copied from class: NeuralNetwork

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.

Overrides:

error in class NeuralNetwork

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.