cameron.freenode.net changed the topic of #mlpack to: http://www.mlpack.org/ -- We don't respond instantly... but we will respond. Give it a few minutes. Or hours. -- Channel logs: http://www.mlpack.org/irc/
zoq has quit [Ping timeout: 264 seconds]
tarun_ml_enthusi has joined #mlpack
< tarun_ml_enthusi>
hello naywhayare ...had been busy for a week..finally free now....so ive installed mlpack and was running simple programs and messing around with armadillo , last time you advised me to stick to an algorithm i know like maybe linear regression and try to fix bugs regarding the same...so i created a trac account and was looking at some of the current active tickets...couldnt find any related to linear regression..
< tarun_ml_enthusi>
can you show me some easy bugs which i can try fixing to get the hang of it?
< naywhayare>
tarun_ml_enthusi: I don't think there are any problems with the linear regression code... so maybe you can find other bugs on Trac that are interesting to you?
< naywhayare>
it's difficult to recommend "easy" bugs because most of them aren't easy...
< tarun_ml_enthusi>
hmm..so how do i suggest i start off? by going through the entire structure of mlpack or by directly diving into a bug and doing some research on it?
< tarun_ml_enthusi>
you*
< naywhayare>
tarun_ml_enthusi: I think either approach can work. I just filed a bug, #376, which I detailed pretty heavily and you might be able to dig into
< naywhayare>
but I don't know if that's a good fit; cosine trees aren't a typical machine learning technique
< naywhayare>
if you aren't afraid to do some reading, maybe it is a good place to start
< naywhayare>
another option is to implement an entirely new algorithm that mlpack doesn't have, or to take an existing algorithm and find a way to make it run faster
< naywhayare>
both of those might be more fun, depending on your interests :)
< naywhayare>
#372 might be interesting and straightforward, too, if you're willing to learn a bit about collaborative filtering
< tarun_ml_enthusi>
yes i did learn a little about collaborative filtering in recommender systems once...ill read up more and try understanding what that bug is about
< tarun_ml_enthusi>
thank you :)
< naywhayare>
cool; if I can explain it better, or if you want tips on how to reproduce the problem, let me know and I can probably give a better description
< tarun_ml_enthusi>
yeah ok ... ill try my luck first and then ask you if i have any doubts :)