fr.lip6.jkernelmachines.kernel.typed

Class DoubleHPolynomial

java.lang.Object

fr.lip6.jkernelmachines.kernel.Kernel<double[]>

fr.lip6.jkernelmachines.kernel.typed.DoubleHPolynomial

All Implemented Interfaces:

java.io.Serializable

public class DoubleHPolynomial
extends Kernel<double[]>

Homogenous Polynomial kernel on double arrays. return (x'y)^d, where d = 2 by default

Author:

David Picard

See Also:

Serialized Form

Field Summary

Fields inherited from class fr.lip6.jkernelmachines.kernel.Kernel

name

Constructor Summary

Constructors

Constructor and Description
DoubleHPolynomial() Default constructor, d = 2
DoubleHPolynomial(int degree) Constructor specifying the degree of the polynomial kernel

Method Summary

Modifier and Type	Method and Description
double	valueOf(double[] t1) kernel similarity to zero
double	valueOf(double[] t1, double[] t2) compute the kernel similarity between two element of input space

Methods inherited from class fr.lip6.jkernelmachines.kernel.Kernel

getKernelMatrix, getKernelMatrixLine, getNormalizedKernelMatrix, normalizedValueOf, setName, toString

Methods inherited from class java.lang.Object

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

Constructor Detail

DoubleHPolynomial

public DoubleHPolynomial()

Default constructor, d = 2

DoubleHPolynomial

public DoubleHPolynomial(int degree)

Constructor specifying the degree of the polynomial kernel

Parameters:

degree - the exponent to which the dot product is raised

Method Detail

valueOf

public double valueOf(double[] t1,
                      double[] t2)

Description copied from class: Kernel

compute the kernel similarity between two element of input space

Specified by:

valueOf in class Kernel<double[]>

Parameters:

t1 - first element

t2 - second element

Returns:

the kernel value

valueOf

public double valueOf(double[] t1)

Description copied from class: Kernel

kernel similarity to zero

Specified by:

valueOf in class Kernel<double[]>

Parameters:

t1 - the element to compute the similarity to itself