07:04 < moksh> Hi rcurtin. Yes I did take a look at those pages, I followed the general instructions, including building from source and implementing some models. I'm also looking at the existing RL models in mlpack. I needed some advice on how to move forward since this is the first time I'm working with mlpack.

07:30 < satyam_2401> By interface you mean pseudo code, right? Could you please show me an example for the interface maybe for the unconstrained case, so that I can get the idea.

10:39 < Atharva> When trying to write tests and pass arguments, why do we have to specify the type as (string) "someargument"?

13:00 < yoloman> Hello everyone.I'm interested in applying to this organisation for Gsoc 2018.I found the idea Reinforcement learning intersting as I have already some knowledge regarding and implemented small projects on it using python.As suggested on the ideas page successfully tested the open AI gym example and read the reinforcement learning code mentioned there.As there are no open issues regarding this topic how should I proceed further in ord

13:03 < zoq> yoloman: Hello there, one idea is to implement a simple RL based on the existing codebase, e.g. Policy Gradients

13:04 < yoloman> ok I will try that.Also do I need to show what I implemented before mentioning it in the proposal to somone?

13:06 < zoq> yoloman: That's not necessary if you could provide a link in your proposal that's just fine.

13:07 < zoq> satyam_2401: Right, http://www.mlpack.org/docs/mlpack-git/doxygen/optimizertutorial.html here is the description of the current interface or you could look at the existing implementations: https://github.com/mlpack/mlpack/tree/master/src/mlpack/core/optimizers.

13:08 < yoloman> Thanks.I will try to implement that as soon as possible.

13:08 < zoq> yoloman: Sounds good, let me know if I should clarify anything.

13:09 < yoloman> Sure.Also is there any other means to contact you apart from irc if i have to clarify something?

13:12 < zoq> yoloman: We like to keep the discussions as open as possible, to give anyone a chance to jump in, so mailing list, IRC or Github would be great

13:13 < yoloman> Sure.I will keep you update about my progress than.Thanks for helping.

13:15 < zoq> yoloman: No problem, here to help :)

15:29 < travis-ci> ShikharJ/mlpack#82 (ResizeLayer - 03864a0 : Shikhar Jaiswal): The build has errored.

15:29 < travis-ci> Change view : https://github.com/ShikharJ/mlpack/compare/812f9e68af30...03864a08fcb2

15:29 < travis-ci> Build details : https://travis-ci.org/ShikharJ/mlpack/builds/344346796

20:05 < satyam_2401> https://www.irccloud.com/pastebin/skfrDH0m/

21:42 < zoq> satyam_2401: Looks good to me, I should note there is an open PR which implements the standard PSO method based on that design. I think this is a really good start to add new features, let me know what you think.

22:00 < marioloko> Hello everybody! My name is Adrian and I am a Google Summer of Code student and I am really interested in some of your ideas :) I come here in order to introduce myself, ask for some advices for getting involved in mlpack and also for asking for advices about what things should I prepare before starting to work.

22:01 < zoq> marioloko: Hello, thanks for getting in touch, www.mlpack.org/involved.html and mlpack.org/gsoc.html should be helpful.

22:03 < marioloko> zoq: Thank you for answer back so fast! I will read it and ask here if I have any doubt :)

22:04 < zoq> marioloko: Sounds good, let me know if I should clarify anything.

22:04 < marioloko> zoq: Perfect! And thank you again :)