00:00 < rcurtin> looks like it is fixed now :)

00:17 < shrit[m]> Good to hear

04:01 < NippunSharmaGitt> Hi all, I saw that we do not have a Gaussian Process Regressor inside mlpack. I think it would be a great tool to have and have opened #2764 for the same. What do you all think about it ?

14:33 < iamarchit123Gitt> what about decision tree/random forest regressor @NippunSharma I have seen it their in scikit-learn couldn't help but wonder.

14:56 < NippunSharmaGitt> @iamarchit123 random forest and decision tree regressor are already implemented inside mlpack

14:58 < NippunSharmaGitt> sorry, the classifiers are implemented but regressors are not

15:14 < rcurtin[m]> it would be great to add decision tree regressors too

15:14 < rcurtin[m]> GP regressors would be nice to add, but at least I am not a GP expert so I don't know that I could do a good job of reviewing the PR

15:14 < iamarchit123Gitt> @rcurtin let me see some implementations on net and give it a try

15:28 < rcurtin[m]> iamarchit123 (Gitter): I'd suggest looking at papers instead of copying other implementations---much better to implement the original equations, instead of reverse-engineering the equations from an implementation and then implementing those

15:29 < rcurtin[m]> it should be possible to implement this functionality using the existing DecisionTree class

15:30 < iamarchit123Gitt> sure @rcurtin let me find some good papers on the topic to implement

15:31 < rcurtin[m]> here's "the big one": https://www.stat.berkeley.edu/~breiman/randomforest2001.pdf

15:32 < iamarchit123Gitt> ill go through the decision tree class of mlpack though i think they might be splitting based on entropy dont know if regression follow similar approach

15:32 < iamarchit123Gitt> > `rcurtin (@ryan:ratml.org)` here's "the big one": https://www.stat.berkeley.edu/~breiman/randomforest2001.pdf

15:32 < iamarchit123Gitt> thanks

15:32 < rcurtin[m]> the existing implementation should be general enough to allow new splitting strategies via templates :)

15:32 < NippunSharmaGitt> > `rcurtin (@ryan:ratml.org)` GP regressors would be nice to add, but at least I am not a GP expert so I don't know that I could do a good job of reviewing the PR

15:32 < NippunSharmaGitt> I am sure that you know a lot about it than me :)

15:33 < rcurtin[m]> Nippun Sharma (Gitter): maybe; in any case, unless we have someone knowledgeable about GP regressors to review the PR, it might not be a great idea to implement it---we could miss some important detail, or implement something that's not the best algorithm, etc.

15:34 < rcurtin[m]> anyway, up to you

15:34 < NippunSharmaGitt> Hmm, I agree. Any other algorithms you can suggest ?

15:35 < rcurtin[m]> not off the top of my head---there are lots of open (and closed stale) issues suggesting new support, although it may be more useful to try and improve existing support in some way; like I said, up to you to find what's interesting to you and then go for it :)

15:38 < NippunSharmaGitt> Okay I will look for some issues and in the process if I come across something new I will ask here.

18:20 < HARSHCHAUHAN[m]> Hii Everyone,

18:20 < HARSHCHAUHAN[m]> I am a newbie in this open-source world and now I would like to contribute in it. Can anyone help me out from where should I can start my journey? I have good holds in ML/DL algos that's why I would like to contribute!!!

18:21 < jeffin143[m]> ryan I marked the boost issue as done and closed it

18:22 < rcurtin[m]> awesome! glad to be done with that :)

18:23 < jeffin143[m]> I am planning to add text rank algorithm for keywords extraction to mlpack

18:24 < jeffin143[m]> :) waiting for the string pr to get merged and then I can start may be

18:24 < rcurtin[m]> yeah, I am excited about the string PRs (I think there are two of them?); it will be great to have that support merged

18:25 < jeffin143[m]> Yes

18:25 < jeffin143[m]> I planned to write some tutorials on strong support that we have

18:25 < jeffin143[m]> And also the visualisation support that we got this summer

18:25 < jeffin143[m]> But there is so much on the plate

18:25 < jeffin143[m]> 😂

18:25 < rcurtin[m]> I know the feeling :)

19:29 < zoq> HARSHCHAUHAN[m]: Hello, the community page / get involved section should be helpful - https://www.mlpack.org/community.html

19:30 < zoq> HARSHCHAUHAN[m]: https://www.mlpack.org/gsoc.html might be helpful as well.

19:43 < HARSHCHAUHAN[m]> zoq Thank you for your reply!! These links help me a lot to understand the community basics and from today onwards I start reviews the issues that are listed in Github repo.

19:44 < HARSHCHAUHAN[m]> (edited) ... start reviews the issues that are listed in Github ... => ... start to review the issues that are listed in the Github ...