weka.classifiers.functions.supportVector
Class Kernel

java.lang.Object
  weka.classifiers.functions.supportVector.Kernel

All Implemented Interfaces:

java.io.Serializable

Direct Known Subclasses:

PolyKernel, RBFKernel

public abstract class Kernel

extends java.lang.Object

implements java.io.Serializable

Abstract kernel. Kernels implementing this class must respect Mercer's condition in order to ensure a correct behaviour of SMOreg.

Version:

$$

Author:

Eibe Frank (eibe@cs.waikato.ac.nz)

See Also:

Serialized Form

Field Summary

(package private) Instances

m_data The dataset

Constructor Summary

Kernel()

Method Summary

abstract void	clean() Frees the memory used by the kernel.
abstract double	eval(int id1, int id2, Instance inst1) Computes the result of the kernel function for two instances.
abstract int	numEvals() Returns the number of kernel evaluation performed.

Methods inherited from class java.lang.Object

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

Field Detail

m_data

Instances m_data

The dataset

Constructor Detail

Kernel

public Kernel()

Method Detail

eval

public abstract double eval(int id1,
                            int id2,
                            Instance inst1)
                     throws java.lang.Exception

Computes the result of the kernel function for two instances. If id1 == -1, eval use inst1 instead of an instance in the dataset.

Parameters:

id1 - the index of the first instance in the dataset

id2 - the index of the second instance in the dataset

Returns:

the result of the kernel function

Throws:

java.lang.Exception

clean

public abstract void clean()

Frees the memory used by the kernel. (Usefull with kernels which use cache.) This function is called when the training is done. i.e. after that, eval will be called with id1 == -1.

numEvals

public abstract int numEvals()

Returns the number of kernel evaluation performed.

Returns:

the number of kernel evaluation performed.