svmhyper.m

this a SVM toolbox,it is very useful for someone who just learn SVM.In order to be undestood easily,

function w = svmhyper(Xtrn,Itrn,alpha)
% SVMHYPER coputes normal vector of linear SVM decision rule.
% w = svmhyper(X,I,Alpha)
% 
% The decision hyperplane is formed by the points x for which 
% equation w'*x + bias = 0 holds.
%
% Input:
%  X [DxM] training patterns.
%  I [1xM] labels of training patterns.
%  Alpha [Mx1] Lagrange multipliers.
%
% Output:
%  w [Dx1] normal vector.
%
% See also SVMCLASS, SVM.
%

% Statistical Pattern Recognition Toolbox, Vojtech Franc, Vaclav Hlavac
% (c) Czech Technical University Prague, http://cmp.felk.cvut.cz
% Modifications
% 19-September-2001, V. Franc, comments changed.

dim=size(Xtrn,1);

Ytrn=3-2*Itrn;

w = (Xtrn.*repmat(Ytrn,dim,1))*alpha';

return;