KNORM |
Computes L2-norm in kernel space.
Synopsis:
kn = knorm(X,Alpha,ker,arg)
Description:
kn = knorm(X,Alpha,ker,arg) computes kn = sqrt(Alpha'*K*Alpha)
where K = kernel(X,ker,arg) is the kernel matrix.
Input:
Alpha [num_data x 1] Expansion coefficients (weights).
X [dim x num_data] Data vectors.
ker [string] Kernel identifier (see "help kernel").
arg [...] Kernel argument.
Ouput:
kn [1x1] Norm in the kernel space.
See also:
KERNEL, KERNELPROJ.
About: Statistical Pattern Recognition Toolbox
(C) 1999-2004, Written by Vojtech Franc and Vaclav Hlavac
Czech Technical University Prague
Faculty of Electrical Engineering
Center for Machine Perception
Modifications:
08-oct-2004, created