[prev in list] [next in list] [prev in thread] [next in thread] 

List:       wekalist
Subject:    [Wekalist] How to fix the number of clusters in clustering
From:       gejianye () gmail ! com
Date:       2020-11-12 1:16:30
Message-ID: 160514379002.1719.13280912981777040307 () sys-mailman-prd ! its ! waikato ! ac ! nz
[Download RAW message or body]

Hi all,

I am trying to guess the best number of clusters by Silhouette coefficient. What I \
did was to set the number of clusters from 2, 3, ..., 10, and for each fixed number \
of clusters, do the clustering and use ClusterEvaluation to calculate Silhouette \
coefficient. The best number of clusters would be the number with the highest \
Silhouette coefficient.

However, what I found was when I set the number of clusters in the WEKA clustering \
algorithms, after clustering, the number of clusters may change to a different \
number. For example, I set the number of clusters as 2, but it may change to 1, 3 or \
other numbers after clustering. The followings are my codes.

			EM model = new EM(); 
			model.setNumClusters(i); // i is the number of clusters, from 2 to 10;
		        model.buildClusterer(data);
			
So, my question is how can I fix the number of clusters in clustering, so that the \
clustering would not try the other numbers of clusters?

Thanks!
JG
_______________________________________________
Wekalist mailing list -- wekalist@list.waikato.ac.nz
Send posts to wekalist@list.waikato.ac.nz
To unsubscribe send an email to wekalist-leave@list.waikato.ac.nz
To subscribe, unsubscribe, etc., visit \
https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz List \
etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html


[prev in list] [next in list] [prev in thread] [next in thread]