covMultiClass.m

function [ K ] = covMultiClass(hyp, para, X1, X2, i)

% Copyright (c) University of Glasgow in UK  - All Rights Reserved
% Author: Li Sun (Kevin) <lisunsir@gmail.com>
% Institute: University of Glasgow
% Details: Kernel Computation for multiple class GP Clasification
% Reference: 
% 1. <Gaussian Process for Machine Learning> 
% 2. <Recognising the Clothing Categories from Free-Configuration using Gaussian-Process-Based Interactive Perception>


kernel = para.kernel;
c = para.c;

if strcmp(kernel,'linear')
    Kci = computeCov(X1, X2, para);
end


if nargin == 4
    Kci = feval(kernel, hyp, X1, X2);
elseif nargin == 5
    Kci = feval(kernel, hyp, X1, X2, i);
else
    disp('ERROR: Too many input variables!');
end


if c == 2
    K = Kci;
elseif c > 2
    Kc = {Kci};
    Kc = repmat(Kc, [c,1]);
    
    K = constructBlockDiag(Kc);
else
    disp('ERROR: c must larger than 2!');
end