On Thu, Feb 13, 2014 at 03:18:00PM -0800, Charles Greenberg wrote: > Sorry to bring up this old thread, but it'd be very helpful (and I imagine, > fairly trivial to implement) an option to assign weights to the data points > used in GMM fitting. It can allow some pretty cool things, like directly > fitting GMMs to a grid where a probability density is assigned to each grid > point. At least, I'd find it cool!
Fitting a GMM to a PDF has no sens. A GMM is a density estimator. It works from data points, not density estimates. Fitting Gaussians on a histogram is another problem that must be tackled with other algorithms. However, it will give a very bad estimator, and I believe that it should not be used. Do you have any other usecases in mind for fitting a GMM to weighted data points? Gaël ------------------------------------------------------------------------------ Android apps run on BlackBerry 10 Introducing the new BlackBerry 10.2.1 Runtime for Android apps. Now with support for Jelly Bean, Bluetooth, Mapview and more. Get your Android app in front of a whole new audience. Start now. http://pubads.g.doubleclick.net/gampad/clk?id=124407151&iu=/4140/ostg.clktrk _______________________________________________ Scikit-learn-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
