clc
clear
load fisheriris
x=meas;
y=species;
mdl=fitcknn(x,y,'NumNeighbors',5,'distance', 'seuclidean','Standardized',1);
c=crossval(mdl,'KFOLD',10);
%KFOLD meaning that for the 150 pieces of info, remove 10 and work on the %rest??
Pred=kfoldPredict(c);
confmat=confusionmat (y, Pred);
Afold=100*(1 kfoldLoss (c,'mode','individual'))
acc=mean(Afold)
run the code with k values from 1 to 10, each time change the distance parameter in the Fitcknn order, draw the results in excel sheet and tell which distance parameter was the best and which k value was the best.