edu.udo.cs.myGP
Class Model

java.lang.Object
  edu.udo.cs.myGP.Model

public class Model
extends java.lang.Object

The learned model.

Version:

$Id: Model.java,v 1.6 2006/09/30 00:05:30 ingomierswa Exp $

Author:

Piotr Kasprzak, Ingo Mierswa

Field Summary

private Jama.Matrix	alpha
private double[][]	basisVectors
private Jama.Matrix	C
private int	d
private int	inputDim
private Jama.Matrix	k Other variables (can be derived from the variables above)
private Kernel	kernel Variables describing the GP (must be saved with the model)
private Jama.Matrix	Q
private boolean	regression

Constructor Summary

private	Model() Constructors
	Model(Kernel kernel, double[][] basisVectors, Jama.Matrix alpha, Jama.Matrix C, Jama.Matrix Q, int d, boolean regression)

Method Summary

double[]	apply(double[][] inputVectors) Apply the model to all input vectors
double	applyToVector(double[] x_new) Apply the model to a (new) input vector x_t+1 in order to get a prediction, which - as a GP-marignal at x_t+1 - is a one-dimensional gaussian distribution with mean m and covariance sigma^2 (2.22, the parameterisation lemma).
double[]	getBasisVector(int i)
double	getBasisVectorValue(int i, int j)
int	getInputDim()
int	getNumberOfBasisVectors()
private static double[]	readDoubleArray(java.io.ObjectInputStream in, int length) Read double[] from an input stream
private static Jama.Matrix	readMatrix(java.io.ObjectInputStream in) Read a matrix from a stream
static Model	readModel(java.io.ObjectInputStream in) Reads the model data from the input stream.
private double	scalarProduct(double[][] x, double[][] y, int d) Compute the (canonical) scalar product between x and y, using only the first d components of the vectors
private void	writeDoubleArray(java.io.ObjectOutputStream out, double[] array) Write double[] into an output stream
private void	writeMatrix(java.io.ObjectOutputStream out, Jama.Matrix matrix) Write a matrix to a stream
void	writeModel(java.io.ObjectOutputStream out) Writes the model data into the output stream.

Methods inherited from class java.lang.Object

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

Field Detail

kernel

private Kernel kernel

Variables describing the GP (must be saved with the model)

basisVectors

private double[][] basisVectors

C

private Jama.Matrix C

alpha

private Jama.Matrix alpha

Q

private Jama.Matrix Q

regression

private boolean regression

k

private Jama.Matrix k

Other variables (can be derived from the variables above)

d

private int d

inputDim

private int inputDim

Constructor Detail

Model

private Model()

Constructors

Model

public Model(Kernel kernel,
             double[][] basisVectors,
             Jama.Matrix alpha,
             Jama.Matrix C,
             Jama.Matrix Q,
             int d,
             boolean regression)

Method Detail

getNumberOfBasisVectors

public int getNumberOfBasisVectors()

getInputDim

public int getInputDim()

getBasisVector

public double[] getBasisVector(int i)

getBasisVectorValue

public double getBasisVectorValue(int i,
                                  int j)

scalarProduct

private double scalarProduct(double[][] x,
                             double[][] y,
                             int d)

Compute the (canonical) scalar product between x and y, using only the first d components of the vectors

applyToVector

public double applyToVector(double[] x_new)

Apply the model to a (new) input vector x_t+1 in order to get a prediction, which - as a GP-marignal at x_t+1 - is a one-dimensional gaussian distribution with mean m and covariance sigma^2 (2.22, the parameterisation lemma). Returns only the function value, the mapping to a classification must be done by the invoking method.

apply

public double[] apply(double[][] inputVectors)
               throws java.lang.Exception

Apply the model to all input vectors

Throws:

java.lang.Exception

readMatrix

private static Jama.Matrix readMatrix(java.io.ObjectInputStream in)
                               throws java.io.IOException

Read a matrix from a stream

Throws:

java.io.IOException

writeMatrix

private void writeMatrix(java.io.ObjectOutputStream out,
                         Jama.Matrix matrix)
                  throws java.io.IOException

Write a matrix to a stream

Throws:

java.io.IOException

readDoubleArray

private static double[] readDoubleArray(java.io.ObjectInputStream in,
                                        int length)
                                 throws java.io.IOException

Read double[] from an input stream

Throws:

java.io.IOException

writeDoubleArray

private void writeDoubleArray(java.io.ObjectOutputStream out,
                              double[] array)
                       throws java.io.IOException

Write double[] into an output stream

Throws:

java.io.IOException

readModel

public static Model readModel(java.io.ObjectInputStream in)
                       throws java.io.IOException

Reads the model data from the input stream.

Throws:

java.io.IOException

writeModel

public void writeModel(java.io.ObjectOutputStream out)
                throws java.io.IOException

Writes the model data into the output stream.

Throws:

java.io.IOException