On Thu, Jan 24, 2013 at 12:34:31AM +0100, Andreas Mueller wrote: > Sorry, custom metrics for K means are not possible at the moment.
Yes, there is a massive difference in amount of work and performance when you try to replace the Euclidean distance. Amongst other problems, the mean is no longer the sum divided by the number of points, but the Frechet mean, which requires solving an optimization problem. Ariel, quite often, I find that use of distances adapted to the unit sphere for clustering is over sold. If you have enough clusters, they do not extend much on the sphere, and thus the sphere is locally equivalent to a plane, and you can use standard Euclidean-based clustering algorithms. If not, you have to pay the price, and it will be expansive. G ------------------------------------------------------------------------------ Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS, MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft MVPs and experts. ON SALE this month only -- learn more at: http://p.sf.net/sfu/learnnow-d2d _______________________________________________ Scikit-learn-general mailing list Scikit-learn-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
