fr.lip6.kernel
Class Kernel<T>

java.lang.Object
  fr.lip6.kernel.Kernel<T>

Type Parameters:

T - Type of data from input space

All Implemented Interfaces:

java.io.Serializable

Direct Known Subclasses:

CMKernel, DoubleGaussChi1, DoubleGaussChi2, DoubleGaussL2, DoubleLinear, FloatGaussChi2, GeneralizedDoubleLinear, GeneralizedSparseDoubleLinear, IndexDoubleGaussL2, IndexDoubleLinear, IntGaussChi1, IntGaussChi2, IntGaussL2, PowerBagKernel, PowerKernel, SimpleSubListKernel, SparseDoubleGaussChi2, SparseDoubleLinear, SparseGeneralizedSparseDoubleLinear, SPDoubleGaussChi2, SubListKernel, SubListMaxKernel, SubListPowerKernel, ThreadedKernel, ThreadedSumKernel, WeightedSumKernel

public abstract class Kernel<T>
extends java.lang.Object
implements java.io.Serializable

Base class for kernels

Author:

dpicard

See Also:

Serialized Form

Field Summary

java.lang.String

name

Constructor Summary

Kernel()

Method Summary

double[][]	getKernelMatrix(java.util.ArrayList<TrainingSample<T>> e) return the Gram Matrix of this kernel computed on given samples
double[][]	getNormalizedKernelMatrix(java.util.ArrayList<TrainingSample<T>> e) return the Gram Matrix of this kernel computed on given samples
double	normalizedValueOf(T t1, T t2) kernel similarity normalized such that k(t1, t1) = 1
void	setName(java.lang.String n)
java.lang.String	toString()
abstract double	valueOf(T t1) kernel similarity to zero
abstract double	valueOf(T t1, T t2) compute the kernel similarity between two element of input space

Methods inherited from class java.lang.Object

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

Field Detail

name

public java.lang.String name

Constructor Detail

Kernel

public Kernel()

Method Detail

valueOf

public abstract double valueOf(T t1,
                               T t2)

compute the kernel similarity between two element of input space

Parameters:

t1 - first element

t2 - second element

Returns:

the kernel value

valueOf

public abstract double valueOf(T t1)

kernel similarity to zero

Parameters:

t1 - the element to compute the similarity to zero

normalizedValueOf

public double normalizedValueOf(T t1,
                                T t2)

kernel similarity normalized such that k(t1, t1) = 1

Parameters:

t1 - first element

t2 - second element

Returns:

normalized similarity

getKernelMatrix

public double[][] getKernelMatrix(java.util.ArrayList<TrainingSample<T>> e)

return the Gram Matrix of this kernel computed on given samples

Parameters:

e -

Returns:

double[][] containing similarities in the order of the e.

getNormalizedKernelMatrix

public double[][] getNormalizedKernelMatrix(java.util.ArrayList<TrainingSample<T>> e)

return the Gram Matrix of this kernel computed on given samples

Parameters:

e -

Returns:

double[][] containing similarities in the order of the e.

setName

public void setName(java.lang.String n)

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object