ChanServ changed the topic of #mlpack to: "mlpack: a fast, flexible machine learning library :: We don't always respond instantly, but we will respond; please be patient :: Logs at http://www.mlpack.org/irc/
jeffin143 has joined #mlpack
jeffin143 has quit [Client Quit]
jeffin143 has joined #mlpack
jeffin143 has quit [Client Quit]
jeffin143 has joined #mlpack
jeffin143 has quit [Client Quit]
jeffin143 has joined #mlpack
mulx10 has joined #mlpack
Toshal has joined #mlpack
Toshal_ has joined #mlpack
Toshal_ has quit [Remote host closed the connection]
mulx10 has quit [Quit: Page closed]
Toshal has quit [Ping timeout: 252 seconds]
mulx10 has joined #mlpack
jeffin143 has quit [Ping timeout: 252 seconds]
xiaohong has joined #mlpack
< xiaohong>
will we have a meeting in 40 minutes?
< rcurtin>
xiaohong: yeah, that is what it is scheduled for :)
< rcurtin>
marcus was going to run it---for me it is very late right now, actually I am surprised I am still awake
< rcurtin>
I will probably fall asleep partway through, hehe
< xiaohong>
Got it, have a good dream.
< lozhnikov>
jeffin143: I think the tokenizer could be quite complicated. It is not necessarily a standard one. I suggest to introduce a template parameter e.g. "Tokenizer" which has e.g. operator () which accepts string or string_view and returns a token.
< lozhnikov>
Actually, we can't implement all possible ways of tokenization. In that case we should introduce a way of custom tokenization.
< Suryo>
rcurtin: yeah, I couldn't login to the last one :/
< Suryo>
So I can't miss this, obviously.
< zoq>
The meeting will be logged.
< rcurtin>
Suryo: yeah, sorry about that. Hangouts was better than Jitsi but there are still some improvements we can make
< zoq>
agreed
saksham189 has joined #mlpack
< Suryo>
I was just unable to connect to it. I was on the road at the time and tried to connect using the link. It only took me to the play store downloads page for the app, whereas I already had the app installed!
< Suryo>
So that's what happened.
< Suryo>
Anyway, less of this rambling right now.
< zoq>
Okay, let's start.
< rcurtin>
:)
< xiaohong>
:)
< zoq>
Hello everyone! I'm Marcus, one of the organization administrators for mlpack. I think I have met all of you (IRC, Guthub or via the mailing list).
< zoq>
The purpose of this meeting is basically to welcome everyone to GSoC, go quickly over how the program works, what the expectations are for students and mentors, and then we can all introduce ourselves and get to know each other a little bit.
< zoq>
As I already said, The channel is logged, so anyone who wants to refer to this later, you can see the logs at http://www.mlpack.org/irc/
< zoq>
Just checked, works.
Toshal_ is now known as Toshal
< zoq>
With all that in mind, congratulations to all of you for being accepted!
< zoq>
This year was once again very competitive. We had about 90 applications for 9 spots; it was difficult to select the best applications from all these, but we did our best and we selected each of you.
< zoq>
Right now, we are in the Community Bonding period, which lasts until May 27th.
< zoq>
The period is a good time to get to know your mentor, the community, poke around the code etc.
< zoq>
Discussion in here (#mlpack) isn't limited to just discussion of mlpack; in previous years, for instance, we've had some nice discussions on music recommendations. Always excited to get some recommendations, so if you have some, please post them here :)
< zoq>
as well as Github and there is also the mlpack-git mailing list, which sends emails for all mlpack-related activity (http://lists.mlpack.org/mailman/listinfo/mlpack-git); the last one is quite noisy :)
< zoq>
Once the community bonding period is over, coding starts on May 27th.
< zoq>
Note the community bonding period being over doesn't mean that we can't still talk about non-mlpack things; we are a friendly community (I think!), so you are free to participate in that community in any way that you like.
< zoq>
During the summer, there are three evaluation periods: from June 24 - 28th, July 22 - 26th, and then the final evaluation from August 26th until September 2nd.
< zoq>
There's no need to be too afraid of the evaluations, we've selected each of you as we clearly think you are capable of doing the project that you proposed.
< zoq>
After GSoC ends on August 26, then we hope that you will hang around, continue to be a part of the community, and then maybe participate in future GSoCs as a student or mentor :)
< zoq>
So, that's the basic timeline.
< zoq>
Before I move on to expectations, any questions, clarifications or comments?
< Suryo>
zoq: for how many years, in all, has mlpack been a part of GSoC?
< Suryo>
And for how many years have you mentored?
< rcurtin>
the timeline sounds good to me :)
< zoq>
If I remember right for 6 years now.
< zoq>
This should be the year number seven.
< Suryo>
Wow, that's wide window of time!
< zoq>
We didn't get in 2015, none of the ML related projects :(
< zoq>
Ok, I will move on.
< xiaohong>
zoq: will be the mlpack-git mailist used more often?
< xiaohong>
sorry for interrupt.
< zoq>
Its basically a Bot that forwards git commit messages :)
< zoq>
If there are any other questions do feel free to interrupt me anytime :)
< zoq>
Next we can talk about student and mentor expectations.
< zoq>
Hopefully, the expectations seem reasonable to anyone and nobody should be apprehensive or worried.
< zoq>
During the summer, students are expected to be working the equivalent of a full-time job (In Germany this is "40 hours a week" but in other countries it's slightly different, so, whatever your definition of 'full-time job' is can apply).
< zoq>
But there is no hard requirement on that, if you'd like to work more some weeks and less some other weeks, that is perfectly okay, so long as the work you've proposed is continuing to get done.
KimSangYeon-DGU_ has joined #mlpack
< zoq>
It's also true that timelines in the proposal tend to be very different than reality.
< zoq>
Sometimes projects go faster than expected, and sometimes they go slower than expected.
< zoq>
If either of these things is happening with your project, you should approach your mentor and work with them to update the timeline accordingly.
< zoq>
Sometimes this means that some of your goals will not fit in the summer. Sometimes it means that you didn't plan enough work and can actually add more.
< zoq>
But it is certainly not a problem if a situation like this happens, so long as you are working with your mentor to ensure that you have a solid plan in place throughout the summer.
< zoq>
Adding on that point, it's really important to maintain regular communications with your mentor.
< zoq>
You should be communicating with your mentor nearly every day that you are working, and they should be kept up-to-date with what you are doing.
< zoq>
Usually it can be really helpful to regularly push your code to your Github fork of the repository, so that your mentor can keep an eye on it and comment.
< zoq>
If you don't keep in touch with your mentor, then we as an organization don't know what you're up to, and this can be a problem at evaluation time. We are supposed to evaluate your progress, and so if we don't know what you have been doing it is difficult to evaluate.
< zoq>
Also, do be sure to keep in touch because disappearing students are a common thing in the entire GSoC program, and generally students who disappear have to be failed since they don't do the work they proposed.
< zoq>
How you communicate with your mentor is up to you, but we'd encourage you to communicate in a public way (mailing list, IRC, github issues).
< zoq>
That way, others can jump in and help.
< zoq>
Students are also expected to provide a weekly update of their project to the community.
< zoq>
In the past, this has been done with a weekly email post, or a post to the mlpack blog: http://mlpack.org/blog/
< zoq>
The blog repository can be found at https://github.com/mlpack/blog/, and we can get each of you push permissions for that repository.
< zoq>
But it is up to you if you want to use a weekly email to the list, or a blog post, or perhaps something else.
< zoq>
We are not picky so long as the community is able to stay up to date.
< zoq>
These weekly updates are nice for others in the community who may not be your mentor or may not even be a part of GSoC to see what is going on inside the program.
< zoq>
The blog (http://mlpack.org/blog/) has lots of examples of weekly updates from previous years, so you could take a look at those to get an idea.
< zoq>
It's perfectly okay for students to take vacation time or be otherwise unavailable (like maybe you have a trip planned already or something).
< zoq>
But so that we don't think that you have disappeared, it's really important that you let your mentor know when you will be unavailable.
< zoq>
Anyway, it's generally expected that students will be respectful towards the community and the community is respectful towards the student!.
< zoq>
I don't think we've ever had any real incidents of people acting disrespectful (correct me if I'm wrong), but if you are ever in a situation where someone's said something to you that's off-putting (even if it's me!),
< zoq>
Please do speak up so that we can resolve the situation in a friendly way.
< zoq>
Before I move on to mentor expectations, let's pause for a minute, in case anyone has any questions or comments?
< Suryo>
Regarding the blog
< Suryo>
Do you have any restrictions on the platforms where we can post, etc.?
< Suryo>
Can it be on medium or bloggr or anything?
< zoq>
Yeah, sure as long as it's not behind a paywall.
< Suryo>
Okay. Thanks.
< rcurtin>
I just counted, I think that we have had 32 GSoC students over the years and this year that will go up to 42 :)
< zoq>
We should post the link on the list and in #mlpack as well.
< rcurtin>
all previous projects have been successes for the students, and I hope we can keep that trend up this year :)
< zoq>
agreed
< mulx10>
Regarding weekly updates, can it include things that are not in proposal?
< zoq>
Yes, the weekly updates should at least cover what you have done over the week.
< mulx10>
Okay thanks
< xiaohong>
If we miss a week due to the vocation
< zoq>
If I remember right Andrew ended always with an interesting quote from a book?
< xiaohong>
we continue our work or
< rcurtin>
zoq: yeah, I always liked those. that was 2016 I think? or was it 2014? I forget
< Suryo>
rcurtin: 42 -- thd meaning of life, the universe, and everything! :D
< rcurtin>
:)
< Suryo>
*the
< zoq>
No need to post an update if you are on vaction, just let us know before.
< xiaohong>
Okay, thanks.
< zoq>
I put some of his recommendations on my list :)
< zoq>
Also if you have any questions during the summer or at any time do feel free to ask!
< Suryo>
'42', it's from the Hitchhiker's Guide to the Galaxy. Just a fun thought.
< zoq>
Moving on to mentor expectations; these are important to point out so that students know what they can expect.
< zoq>
Mentors are expected to work with their students to find times that both they and the student are regularly available to communicate and work with each other.
< zoq>
Mentors should also be willing to help debug code problems and help students understand theory or papers as needed.
< zoq>
Definitely the mentor is not expected to do the work of the project, but, they should be there to help when they are needed.
< zoq>
Just like for students, it's important that the mentor notifies their students (and probably others) if they will be unavailable during the summer period due to a vacation or similar.
< zoq>
In general, mentors should aim to communicate with their student in public so the community can see the progress of the project.
< zoq>
But like I said earlier this is not a firm requirement, and whatever works best for a given student-mentor combination should be used.
< zoq>
Mentors are required to submit the two midterm evaluations and final evaluation for the student, Google is pretty firm about the deadline.
< zoq>
Google sends emails basically every day when the evaluations haven't been done :)
< zoq>
Mentors should also track the progress of their student's project and make sure that the project is proceeding according to the timeline.
< zoq>
If this is not happening (for whatever reason), they should work with the student to establish a more realistic timeline and set of goals.
< zoq>
Mentors shouldn't choose to fail students without any warning, in any situation like this, the student should be made adequately aware of what the expectations for a passing evaluation are.
< zoq>
So, that does it for the mentor expectations.
< zoq>
The last thing is, to set up the organization properly on Github, I'll need to get everyone's Github accounts.
< zoq>
I guess we already have all of these, or, at least, the ones each of you have used to contribute to mlpack in the past.
< zoq>
But if you want to use a different one for some reason, do let me know.
< zoq>
Any questions about expectations, communications, process, etc.?
< rcurtin>
to me it all sounds good, but I have been through it a few times :)
xiaohong has quit [Ping timeout: 256 seconds]
< rcurtin>
(also I am surprised I have not fallen asleep yet!)
< zoq>
yeah, right :)
< zoq>
2am?
< zoq>
Ok, moving on.
< zoq>
The community has grown a lot, in 2011 it was just Ryan and a couple of undergraduates; working in the lab, and now there are over 140 contributors!
< zoq>
rcurtin: Correct me if I'm wrong, if you are still awake :)
< zoq>
Really cool to watch this growth over the years.
< zoq>
We actually have contributors from every continent except Antarctica, really cool fact :)
< zoq>
Lots of countries from around the world, including some from North Korea.
< rcurtin>
zoq: yeah, that's right, actually it's me who's off---I didn't realize it was 140, I thought it was just above 100. I need to stay updated :)
< zoq>
That reminds me of something; have to send them a message to reactivate my account :)
< zoq>
As Ryan already pointed out, I think, over the years, we have had about 32 GSoC students.
< zoq>
So, that's a quick overview of the mlpack community.
< zoq>
I am sure there are lots of things I have left out.
< zoq>
Really excited about the future.
xiaohong has joined #mlpack
< zoq>
So moving on to the last part, introductions, so we can all get to know each other a bit :)
< jeffin143>
Zoq : "to set up the organization properly on github" means.?
< zoq>
Or are there any questions?
< zoq>
Get you the permissions to write to the wiki, push to the blog, restart the jenkins build.
< jeffin143>
Also ,have always had this question up my mind, we have all those automated binding and markdowns and others, but still I have see less users using mlpack, why don't we have much hype as other libraries have. Or may be we have but I don't know.
< jeffin143>
This was my first gsoc attempt, and that is sole reason I came to know about mlpack, glanced over the project at GSOC program
< xiaohong>
Do we have the permission to restart the jenkins? Is there any risk?
< jeffin143>
Else I wouldn't have encountered it*
< rcurtin>
jeffin143: I can try to answer these questions a little bit---C++ is less popular than other languages for machine learning, and the automatic bindings are still pretty new
< zoq>
I would say, at a conference it's easy to find someone who knows mlpack :)
< jeffin143>
Rcurtin : speed should win over popularity..??
< rcurtin>
jeffin143: and yeah, at conferences I tend to find people who know about it. I think its popularity will continue to increase as time goes on, especially with good bindings to other languages
< rcurtin>
jeffin143: exactly, speed is a nice selling point, but it has to be advertised well :) finding the time to do that is tricky (we don't have a marketing department like Google or Microsoft :))
< zoq>
xiaohong: In this case restart means you just post a comment "@mlpack-jenkins test this please" and the PR get's queued again.
< Suryo>
I'd like to pitch in: my ML teacher at my university knew about armadillo and mlpack and stuff. So that kind of reinforces rcurtin's point. It really depends on what your focus is, and what you're expected to do.
< Suryo>
For instance, scikit learn's backend is written in Cython, which is closer to C++ than Python.
< rcurtin>
Suryo: great to hear, your ML teacher sounds like a cool guy :)
< xiaohong>
Oh, I get it. I am afraid I will ruin someone CI test. lol
< Suryo>
And if I'm not mistaken, that's what makes it faster than native Python implementations.
< zoq>
xiaohong: Nah, you can't do anything wrong here.
< Suryo>
And if I'm right, even numpy's backend is in C with BLAS bindings.
KimSangYeon-DGU_ has quit [Ping timeout: 256 seconds]
< jeffin143>
C++ is overpowering :)
< rcurtin>
Suryo: yeah, numpy, like Armadillo, is basically just a wrapper around LAPACK and BLAS (but without the template metaprogramming parts :))
< zoq>
Looks like there is also a lot of interest in mlpack from the windows platform.
< Suryo>
So what I'm trying to say is while the 'popular' libraries are mostly pythonic, the real stuff happens in languages like C and C++.
< xiaohong>
I will be carefully. :)
< Suryo>
Those are my two cents :)
< zoq>
Okay, moving on to introductions, let me start.
< Suryo>
rcurtin: yeah, he's a really cool guy. He's my research guide, too.
< zoq>
I'm Marcus, I'm located in Germany, Berlin, currently finishing my PhD.
KimSangYeon-DGU_ has joined #mlpack
< zoq>
In my free time, I try to go paragliding and or rock climbing. Always wanted to try a wingsuit, might be the perfect combination :)
< zoq>
Also I really like to watch movies.
< zoq>
The last one I was "First Man".
< zoq>
btw. does anybody watch Game of Thrones, I guess since the show is so popular the chances are high :)
< Suryo>
Yeah I do
< xiaohong>
Yes, I do.
< zoq>
Excited for the final episode?
< Suryo>
Hell yeah
< Toshal>
Zoq: yes indeed.
< jeffin143>
Got fans :)
< Suryo>
I have mixed feelings about the previous episode.
< xiaohong>
But it is a little dispointint about the last episode.
< Suryo>
Wait. WAIT.
< Suryo>
Here's a fact.
< Suryo>
The opening scene wirh Varys
< Suryo>
He was enquiring about Denarys having her dinner, right?
< zoq>
Right
< Suryo>
And he said they'll 'try again'.
< Suryo>
Did you guys realize that he was trying to poison her?
< Suryo>
Let that sink in.
< Suryo>
:D
< zoq>
yes, not sure we will see that, I think Jon will stab Dany :)
< Suryo>
That's an interesting theory haha
< zoq>
Bran will be on the Throne at the end.
< Suryo>
Seems like it.
< zoq>
Anyway, that is my introduction. Everyone else should feel free to introduce themselves in whatever order they like :)
< zoq>
We can't go around a circle or anything since we are all on IRC; I guess we could go alphabetically or something :)
< rcurtin>
I can go next but I only say that because I am about to fall asleep :)
< zoq>
ah right :)
< rcurtin>
I'm Ryan! I'm from Atlanta, and I'm very tired today :)
< rcurtin>
I did my PhD at Georgia Tech, and since then did some time at Symantec and now am at a startup called RelationalAI
< rcurtin>
in my free time I like to race go-karts, and earlier today I was at a national park in South Carolina seeing 'synchronized fireflies' (which I didn't know existed)
< rcurtin>
they all blink at the same time... it was really cool
< rcurtin>
:)
< zoq>
do you have a photo by any chance?
< Suryo>
:O
< rcurtin>
I don't, I think it would be really hard to capture it on camera. they are really dim; but I can get a link, hang on...
< rcurtin>
that's it for me, if someone else wants to introduce themselves they should go ahead :)
< jeffin143>
I would like to do so
< zoq>
rcurtin: When is the all day long race event?
< zoq>
jeffin143: Great :)
< rcurtin>
zoq: not sure when the next 24 hour race is, on Sunday I will just do a short 2 hour race :)
< jeffin143>
I am jeffin sam , from India, to be specific from madhya Pradesh .
< zoq>
rcurtin: Ah, I got that wrong.
< zoq>
rcurtin: Sunday, time to get some sleep :)
< jeffin143>
I am obsessed with c++ , and now with mlpack too. I have always looked to get the output by hook or crook ( being a competitive programmer) , but over the past month coding at mlpack taught me that getting output is not the ultimate aim, how you reach their is important.
< jeffin143>
I have learned a lot in this 1 month, bits of cmake, boost( since we aren't allo
< jeffin143>
Allowed to use boost in online platforms)
< jeffin143>
And other bo
< jeffin143>
Bits of c++*
< jeffin143>
Sorry for the typo 😅
< zoq>
No worries :)
< jeffin143>
I like to watch stand-up comedy in free time* ,. Russell peters being one of my favorite, and some of Indians too
< jeffin143>
Also these days compiling mlpack is my favorite pass time :-p , take 2 hours on my system since lower end
< jeffin143>
Would like to bring it down* ( going through some articles)
< zoq>
I really like Hasan Minhaj, the "White House Correspondents Dinner" is hilarious.
< zoq>
jeffin143: I'll think I can help you there :)
< jeffin143>
Ahaa , I have seen it :-p
< jeffin143>
Zoq : surely , have already gone through that open issue , would talk about it briefly later.
< zoq>
Sounds good.
< zoq>
Thanks for the introduction.
< jeffin143>
So that's it from my side , would like to know about others too
< mulx10>
Okay, I'll like to go about next.
< mulx10>
I am Mehul, an undergraduate at IIT Kharagpur, I like playing football. I usually watch horror, kind of weird but I like it. And I am not really a GOT fan :p
< mulx10>
I got peer pressure here as well :p
< zoq>
The last horror movie I saw was cabin in the woods, not sure that counts as horror.
< Suryo>
zoq: Loll XD
< Suryo>
I used to watch a lot of horror when I was in my undergrad
< Suryo>
One movie that I really liked was The Atticus Institute.
< Suryo>
Thought I'd begin a blog and stuff with my roommate :') good times
< mulx10>
zoq: lol.
< zoq>
Is the blog still online?
< Suryo>
zoq: it technically counts as horror until the penultimate scene and you know what I'm talking about.
< Suryo>
No, we never began it XD
< Suryo>
But we had watched many horror movies.
< zoq>
Suryo: :)
< zoq>
Just opened the Trailer for "The Atticus Institute" ...
< mulx10>
I don't know whether it should be called horror or horrible xD
< Suryo>
mulx10: you mean Cabin in the Woods?
< mulx10>
but moslty i watch horrible movies
< mulx10>
Suryo: in general
< Toshal>
Okay, I would like to go next.
< zoq>
Toshal: Please go ahead.
< Toshal>
Hi, I am Toshal Agrawal an undergraduate student in PICT, Pune, India.
< Toshal>
In my free time I like travelling, reading and playing soccer.
< Toshal>
I also like watching movies especially hilarious ones. :)
< KimSangYeon-DGU_>
Okay, I'd like to go next introduction.
< zoq>
Nice contrast, moving from horror to comedy.
< KimSangYeon-DGU_>
Oops, sorry
< zoq>
Toshal: Are you reading anything right now?
vivekp has joined #mlpack
< Toshal>
zoq: Yeah, I am reading The monk who sold his own ferrari.
< KimSangYeon-DGU_>
Can I introduce myself??
< zoq>
Currently, I'm reading: the "Lunar Rover Manual" :)
< zoq>
KimSangYeon: Yes, sure, please :)
< KimSangYeon-DGU_>
Hi, my name is Sangyeon Kim, I'm from Seoul, Korea
< KimSangYeon-DGU_>
I'm an undergraduate student, studying Computer Science and Engineering at Dongguk University.
< KimSangYeon-DGU_>
In the meantime, I've contributed to mlapck and learned a lot from members of mlpack (of course, I'm a member too, XD)
< KimSangYeon-DGU_>
I think it's lucky to have an opportunity to work with you in the summer.
< KimSangYeon-DGU_>
Actually, I've been thinking about what field I'm interested in and I should major in
< KimSangYeon-DGU_>
By contributing to mlpack, I was sure about my future and I want to grow up my dream with mlpack
< KimSangYeon-DGU_>
And I like to go short-trip whenever my mind gets stuck with something :)
< KimSangYeon-DGU_>
(mountain, river, etc..)
< KimSangYeon-DGU_>
Let's make a great summer!! That's it :)
< saksham189>
I am Saksham from Delhi, India. I am currently pursuing bachelor’s in Computer Science at NSIT, Delhi.
< saksham189>
In the past I have lived in Iran, France and China (in that order). I completed high school in China.
< xiaohong>
China?
< saksham189>
Now my parents have moved to Vietnam and I am studying in India.
< xiaohong>
Which city? lol
< saksham189>
xiaohong: yes, I went to International School of Beijing
< xiaohong>
Wow, great.
< zoq>
saksham189: Interesting, do you have a favorite place?
< xiaohong>
Please continue you intro.
< saksham189>
zoq: no, I don't have to a favourite place to live. But I would like to visit Bali.
< saksham189>
In my free time I like watching movies (I recently saw John Wick 3) and reading books.
< Toshal>
saksham189: That sounds great.
< saksham189>
I am reading flow by mihaly csikszentmihalyi.
< zoq>
Ah action genre, did you liked John Wick 3?
< xiaohong>
Do you finish your intro?
< saksham189>
yup, That's it!
< xiaohong>
I'll like to go next. :)
< xiaohong>
I'm Xiaohong! I'm from Wuhan, China. Currently I am studying in Wuhan universisty.
< zoq>
I'll put "flow" on my list.
< saksham189>
zoq: yes i thought it was great! better than the previous ones
< zoq>
xiaohong: Sorry, please go ahead.
< xiaohong>
no problem.
< xiaohong>
In my free time, I like to swimming, cooking. Swimming is really a good way to relax the body.
< xiaohong>
Also I am fans of Game of Thrones. :)
< Suryo>
Which house?
< xiaohong>
That's all for my intro.
< Suryo>
Not sure if there are too many choices left now, though. Haha.
< zoq>
xiaohong: Thanks
< zoq>
Suryo: Targaryen?
< xiaohong>
It seems that most of us like movies.
< Suryo>
No, Starks forever!!!
< zoq>
xiaohong: yeah, I guess I will update my have to watch list
< Suryo>
Am I the only one left for the intro?
< Suryo>
If there's anyone else, please do introduce yourself before me.
< zoq>
I think so.
< Suryo>
Okay
< Suryo>
I'll wait
< zoq>
I think there is none else, unless I missed someone :)
< Suryo>
Okay. Then I guess I'll introduce myself.
< Suryo>
My name is Suryoday Basak. I'm not a regular GSoC participant -- I was extremely enthusiastic for the program and I suppose I had a good plan, but the problem was that I'm in the US on a student visa and due to job-type restrictions, I could not participate. I thank the organization for keeping me on and I'll be working towards making good contributions.
< Suryo>
I'm going to be working on implementing different variants of particle swarm optimization for ensmallen.
< Suryo>
I have a background in research related to ML methods and applications from my undergrad years in India. Currently, I'm a master's student in the University of Texas at Arlington. I'm interested in pursuing research and my PhD in ML theory and it's intersection with chaos theory and dynamical systems.
< Suryo>
I have a lot of interests. In my teenage, I loved to play basketball and rugby. Since my childhood, I've been a musician - and I'm trained a bit in Western classical as well as Hindustani. Of late, I've been interested in working with djent and Indian sounds for metal. Apart from that, I like to travel and go on adventures - during the video meeting, I was backpacking in San Antonio, and I loved it. I like to read, write and compose
< Suryo>
Also, I like horror. And my favourites would be The Conjuring, Atticus Institute, Insidious, and one of the more recent ones of the Exorcist Series, I think it came out around 2005. Don't remember the exact title.
< Suryo>
In a nutshell, that's it.
< mulx10>
Atticus Institute is really nice :)
< Suryo>
I should also mention that A Wednesday and Dil Chahata Hai are on my list of favorite movies. They're both Bollywood movies.
< Suryo>
mulx10: yeah, it was quite innovative! I liked the style, overall.
< zoq>
Will put both on my list :)
< Suryo>
Be sure to watch City of God.
< Suryo>
Don't watch a dubbed version, of course.
< Suryo>
That's honestly one of the best movies ever...
< zoq>
Didn't they win an oscar?
< Suryo>
I don't think that they did.
< Suryo>
It's unlikely with a movie like that, but it's one of those things, with me at the very least, that struck a chord.
< zoq>
The cover looks familiar. Will move that up :)
< Suryo>
Awesome! :D
< zoq>
Okay, thanks everyone; I guess if there are no further questions we can all get some sleep or get something to eat.
< zoq>
But as I said please feel free to ask questions here, on the mailing list or on github. Happy to clarify anything further.
< Suryo>
Okay! Thanks for the meeting. I'm glad that we had this opportunity to interact.
< KimSangYeon-DGU_>
Agreed, thanks for the meeting!! :)
< xiaohong>
Thanks for the meeting to know each other.
< jeffin143>
Agreed
< ShikharJ>
I guess I'm a bit late for the meeting. Thanks for running this up zoq !
< chandramouli_r>
zoq: what should be done when there is an empty input or if the decomposition fails. Giving fatal outputs gives logic errors while running test cases against it for multiplying matrices of dimensions 0x0 and other nxm. So how to handle these exceptions if there is an empty input or if decomposition fails. What should be the proper way to stop running the code and log out of this.
< rcurtin>
sorry everyone, I was trying to stay awake for introductions but I fell asleep :(
sreenik has joined #mlpack
< sreenik>
Sorry everyone, I recently installed archlinux and it was driving me crazy and I eventually dozed off on the table itself
< sreenik>
Don't know if anyone is around, but I'd like to introduce myself as everyone is done I suppose
< sreenik>
I am Sreenik Seal, a second year engineering undergrad studying electronics at IIEST Shibpur, India.
< sreenik>
Apart from coding interests I have interests in robotics as well. I am not a very big fan of movies (but the ones Suryo mentioned are on my list of great movies too). I like to play computer games instead. I play games across all genres starting from Call of Duty to Age of Empires to NFS and everything in between
< sreenik>
I am also an amateur table tennis player. I like inventing strange serving techniques (and rarely get to work on my gameplay). I have another passtime; I like to practise Graphology (handwriting analysis). It's a very effective way for introspection and also a very fun thing to do. Nowadays requests (from my friends) have decreased so it is generally not more than one handwriting sample a month.
sreenik has quit [Ping timeout: 256 seconds]
< rcurtin>
sreenik: very cool about graphology :)
sreenik has joined #mlpack
< sreenik>
Yeah, thanks :)
< zoq>
agreed, I guess the first time someone mentioned Graphology as a hobby
< zoq>
chandramouli_r: One idea might be to just fail and show the input, not sure there is any need to handle an empty user input, if the decomposition fails anyway.
< chandramouli_r>
but there is some other problem while running travis-ci build. Should I post the log here. I am not able to understand where exactly the problem is in the code. I have tried reading all the source files included in it but no clue.