Comments on: k-means and EM clustering algorithms http://www.sematopia.com/2006/04/k-means-and-em-clustering-algorithms/ Thu, 09 Sep 2010 05:36:22 -0400 http://wordpress.org/?v=2.8.4 hourly 1 By: Preeti http://www.sematopia.com/2006/04/k-means-and-em-clustering-algorithms/comment-page-1/#comment-172519 Preeti Mon, 09 Mar 2009 07:58:18 +0000 http://www.sematopia.com/?p=72#comment-172519 Hi, I would like to know how kmeans will be useful in case of incremental clustering. If the information is dynamically changing, then how kmeans or nearest neighbor will handle editing correct cluster. How to handle unlabelled data? if to begin with we have 90% unlabelled data and 10% labelled then how the algorithm looks like? if this ration need to change, 80% and 20% then how will be the said algorithm change? kindly guide in this regards. Preeti. Hi, I would like to know how kmeans will be useful in case of incremental clustering. If the information is dynamically changing, then how kmeans or nearest neighbor will handle editing correct cluster. How to handle unlabelled data? if to begin with we have 90% unlabelled data and 10% labelled then how the algorithm looks like? if this ration need to change, 80% and 20% then how will be the said algorithm change? kindly guide in this regards.

Preeti.

]]> By: Ravi http://www.sematopia.com/2006/04/k-means-and-em-clustering-algorithms/comment-page-1/#comment-22001 Ravi Thu, 24 May 2007 16:21:12 +0000 http://www.sematopia.com/?p=72#comment-22001 Hi George Thanks for the excellent work! Just wondering how Kmeans can be extended to more then 2 dimensions. Thanks Ravi Hi George
Thanks for the excellent work!
Just wondering how Kmeans can be extended to more then 2 dimensions.

Thanks

Ravi

]]> By: George A. Papayiannis http://www.sematopia.com/2006/04/k-means-and-em-clustering-algorithms/comment-page-1/#comment-6018 George A. Papayiannis Fri, 22 Sep 2006 22:51:33 +0000 http://www.sematopia.com/?p=72#comment-6018 Hi Leilane, The code is licensed uder <a href='http://www.gnu.org/licenses/gpl.txt' rel="nofollow">GPL v2</a>. So go ahead, as long as your confrom to that license. On a side note, as I said in the post I wrote them both in one night. The EM version surely has problems, k-means is probably right... Good luck, George Hi Leilane,

The code is licensed uder GPL v2.
So go ahead, as long as your confrom to that license.

On a side note, as I said in the post I wrote them both in one night.
The EM version surely has problems, k-means is probably right…

Good luck,
George

]]> By: Leilane Castro http://www.sematopia.com/2006/04/k-means-and-em-clustering-algorithms/comment-page-1/#comment-6010 Leilane Castro Fri, 22 Sep 2006 19:16:33 +0000 http://www.sematopia.com/?p=72#comment-6010 Hi! I'm working in a research project about video segmentation and like to know if I can use your source code in this. If I change something I'll let you know, sure. And, in case you say yes, tell me how to thanks you properly in my publication. Thanks for your attention. Hi! I’m working in a research project about video segmentation and like to know if I can use your source code in this. If I change something I’ll let you know, sure. And, in case you say yes, tell me how to thanks you properly in my publication.

Thanks for your attention.

]]>