10:41 < alexsc> yeah, I’ll be using it intensely in the next days.. great toolbox. first thing I can report is about the number of doxygen warnings because of the outdated @param documentation :). so many

10:42 < alexsc> I can fix that on the parts I’ll be using and issue pull requests

11:02 < sachu> hii , want to contribute , need help

11:39 < zoq> sachu: Hello there, http://www.mlpack.org/involved.html might be helpful.

11:51 < alexscc> one question: is it implemented somewhere, for KNN, the selection of the “top ranking class among the best K”? I can code it myself but I wonder if it’s already in the framework

15:16 < rcurtin> alexscc: actually that is not implemented; the original lab group that built the KNN code was more interested in the KNN problem but not the KNN classification problem

15:17 < rcurtin> there are two ways to implement it; the first is to use the existing KNN class and then argmax over the results

15:17 < rcurtin> the second is to actually implement a dual-tree traversal that can have some additional prunes to give additional speedup for classification

15:18 < rcurtin> definitely the first approach is the easiest for now, but I am hoping that the second approach can be implemented eventually (however I haven't had any chance to do that)

18:47 < jjmachan> hai guys

18:49 < jjmachan> I'd like to contribute to this project, I've finished compiling the project.

18:49 < jjmachan> but i'm just learning ml so what would be the best areas where I should focus?

18:51 < zoq> jjmachan: Hello there, that depends on your interests. You could go over the github issues and maybe you find something.

18:52 < zoq> Maybe you like to contribute something new, it's up to you.

18:57 < rcurtin> jjmachan: I opened an issue recently that might be good for someone who wants to learn a little bit of machine learning and also learn about mlpack: https://github.com/mlpack/mlpack/issues/1152

18:57 < rcurtin> maybe you will find that interesting, maybe not---there are lots of other issues too if not :)

19:35 < jjmachan> @zoq thnks :)

19:36 < jjmachan> @rcurtin checked out that issue .... i don't know if I can but I'll try my best!

19:41 < rcurtin> sure, sounds good, hope the pointers are helpful

19:42 < jjmachan> actually its the best written issue I've seen..... really helpful for someone like me thanks a lot @rcurtin

19:49 < rcurtin> good to hear, I will try to be sure to write more issues like that in the future