jml.kernel
Class Kernel

java.lang.Object
  extended by jml.kernel.Kernel

public class Kernel
extends java.lang.Object

Java implementation of commonly used kernel functions.

Version:
1.0 Mar. 29th, 2013
Author:
Mingjie Qian

Constructor Summary
Kernel()
           
 
Method Summary
static org.apache.commons.math.linear.RealMatrix calcKernel(java.lang.String kernelType, double kernelParam, org.apache.commons.math.linear.RealMatrix X)
          Computes Gram matrix of a specified kernel.
static org.apache.commons.math.linear.RealMatrix calcKernel(java.lang.String kernelType, double kernelParam, org.apache.commons.math.linear.RealMatrix X1, org.apache.commons.math.linear.RealMatrix X2)
          Computes Gram matrix of a specified kernel.
static void main(java.lang.String[] args)
           
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

Kernel

public Kernel()
Method Detail

main

public static void main(java.lang.String[] args)
Parameters:
args -

calcKernel

public static org.apache.commons.math.linear.RealMatrix calcKernel(java.lang.String kernelType,
                                                                   double kernelParam,
                                                                   org.apache.commons.math.linear.RealMatrix X)
Computes Gram matrix of a specified kernel. Given a data matrix X (d x n), it returns Gram matrix K (n x n).

Parameters:
kernelType - 'linear' | 'poly' | 'rbf' | 'cosine'
kernelParam - -- | degree | sigma | --
X - a matrix
Returns:
Gram matrix (n x n)

calcKernel

public static org.apache.commons.math.linear.RealMatrix calcKernel(java.lang.String kernelType,
                                                                   double kernelParam,
                                                                   org.apache.commons.math.linear.RealMatrix X1,
                                                                   org.apache.commons.math.linear.RealMatrix X2)
Computes Gram matrix of a specified kernel. Given two data matrices X1 (d x n1), X2 (d x n2), it returns Gram matrix K (n1 x n2).

Parameters:
kernelType - 'linear' | 'poly' | 'rbf' | 'cosine'
kernelParam - -- | degree | sigma | --
X1 - a matrix
X2 - a matrix
Returns:
Gram matrix (n1 x n2)