kmeans
in [Matlab]
|
From: Ozge Taskan on 31 May 2010 04:48 Hi , I try to cluster my data which has dimensions (532,3). I have to use k means algorithm. In general, my problem is an assignment problem. I try to assign my clusters to 4 different modes. I started with k=4. So it is not difficult to look at each assignment possibility since it is just 4^4. Now I have to increase my cluster number, however my mode number will be 4 again. I thought that I need to know how I can find the optimal number of clusters. I could not find any algorithm to determine the optimal number of clusters. Is there any algorithm that I can use with Matlab? thank you very much. Ozge
From: Oleg Komarov on 31 May 2010 05:57 "Ozge Taskan" > Hi , > > I try to cluster my data which has dimensions (532,3). I have to use k means algorithm. In general, my problem is an assignment problem. I try to assign my clusters to 4 different modes. I started with k=4. So it is not difficult to look at each assignment possibility since it is just 4^4. Now I have to increase my cluster number, however my mode number will be 4 again. I thought that I need to know how I can find the optimal number of clusters. I could not find any algorithm to determine the optimal number of clusters. Is there any algorithm that I can use with Matlab? > > thank you very much. > > Ozge You're trying to find a solution that has no definite answer. Check these links: http://www.rd.dnc.ac.jp/~tunenori/doc/487-053.pdf http://home.dei.polimi.it/matteucc/Clustering/tutorial_html/kmeans.html http://www.csse.monash.edu.au/~roset/papers/cal99.pdf Oleg
|
Pages: 1 Prev: removing gamma correction from a PNG image Next: Dicomwrite |