Dear all, I am still not comfortable with k-means clustering in classification. Especially while rerunning with the code
eva = evalclusters(params,'kmeans','silhouette','KList',[2:20]);
It gives me different cases. Also while using separate code like
[idx,cent,sumdist, meansilh] =kcluster('param2W_sac.cld', optkn, s);
there is different answers. I think I am concemptually wrong in using rng-function. Is there some one to help me in this manner. It will be great is there availability to have online tutor in such cases.
Best Answer