On 11/30/2013 12:23 PM, Lars Buitinck wrote: > 2013/11/30 Kyle Kastner <[email protected]>: >> Does anyone have a consolidated list of API issues with HMM? I have been >> working with them recently and would be willing to try and work it into >> better shape, unless core devs would rather have it removed. > The API of the HMMs is fundamentally different from all other > estimators: their fit method takes a *list of* sequences, each a > feature matrix. That means HMMs don't play with pipelines or anything > else in scikit-learn, unless we somehow hack support for this format > into everything else. > > If anything, I would like to remove MultinomialHMM from the next > release. As I said earlier, it's not only a bad fit for scikit-learn, > but also useless for all for a very small set of tasks. > I guess remove means deprecate, right? I am +1 but we should definitely find a place for the code. Worse case it will be a repo with containing just the HMM.
Cheers, Andy ------------------------------------------------------------------------------ Rapidly troubleshoot problems before they affect your business. Most IT organizations don't have a clear picture of how application performance affects their revenue. With AppDynamics, you get 100% visibility into your Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro! http://pubads.g.doubleclick.net/gampad/clk?id=84349351&iu=/4140/ostg.clktrk _______________________________________________ Scikit-learn-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
