Classification matlab code - Quadratic Discriminant Analysis

function [ errorrate,ind,ycom ] = QDAg( Xtest,Xtrain,Ytest,Ytrain,num ) %QDA for data with the seperated Gaussian distribution %num - the number of classes ycom=zeros(length(Ytest),1); sigmainv=cell(1,num); logdet=zeros(1,num); mu=cell(1,num); logpi=zeros(1,num); for i=1:num [sigmainv{i},logdet(i),mu{i},logpi(i)]=genpara(Xtrain,Ytrain,i); end for i=1:length(Ytest) x=Xtest(i,:); delta=zeros(1,num); for j=1:num delta(j)=-1/2*logdet(j)-1/2*(x-mu{j})*sigmainv{j}*(x-mu{j})'+logpi(j); end [~,ycom(i)]=max(delta); end [errorrate,ind]=rate(ycom,Ytest); end