svmsim.m

SVC for Neural Networks (Simon Haykin) homework 6.14

function Yd = svmSim(svm,Xt)

X = svm.x;
Y = svm.y;
a = svm.a;

epsilon = 1e-8;                  
i_sv = find(abs(a)>epsilon) ;     

tmp = a.*Y*kernel(X,X(:,i_sv))  % optimal hyperplane      
b = 1./Y(i_sv)-tmp
b = mean(b)
tmp = a.*Y*kernel(X,Xt);
Yd = sign(tmp+b);  % f(x)=sgn(W'K+b)