--- Log opened Wed Feb 01 00:00:19 2012 | ||
-!- blackburn [~qdrgsm@188.122.228.60] has left #shogun [] | 00:27 | |
-!- naywhayare [~ryan@spoon.lugatgt.org] has quit [Ping timeout: 240 seconds] | 04:13 | |
-!- naywhayare [~ryan@spoon.lugatgt.org] has joined #shogun | 04:14 | |
-!- dfrx [~f-x@inet-hqmc05-o.oracle.com] has joined #shogun | 06:04 | |
CIA-11 | shogun: Soeren Sonnenburg master * raaf479e / (src/shogun/features/Labels.h src/shogun/regression/KRR.h): fix minor documentation typos - http://git.io/u_tnEg | 09:04 |
---|---|---|
-!- sonne|work [~sonnenbu@194.78.35.195] has left #shogun [] | 10:51 | |
-!- dfrx [~f-x@inet-hqmc05-o.oracle.com] has left #shogun [] | 11:37 | |
-!- blackburn [~qdrgsm@188.122.228.60] has joined #shogun | 13:58 | |
-!- Ram108 [~amma@14.96.92.123] has joined #shogun | 15:25 | |
-!- Ram108 [~amma@14.96.92.123] has quit [Ping timeout: 252 seconds] | 15:40 | |
-!- Ram108 [~amma@14.96.92.123] has joined #shogun | 15:50 | |
-!- blackburn [~qdrgsm@188.122.228.60] has quit [Quit: Leaving.] | 16:09 | |
-!- blackburn [~qdrgsm@188.122.228.60] has joined #shogun | 16:19 | |
blackburn | sonney2k: is it true submission could take an year? :) | 17:20 |
-!- Ram108 [~amma@14.96.92.123] has quit [Ping timeout: 252 seconds] | 17:32 | |
-!- Ram108 [~amma@14.96.158.159] has joined #shogun | 17:45 | |
-!- blackburn [~qdrgsm@188.122.228.60] has quit [Ping timeout: 240 seconds] | 18:08 | |
-!- blackburn [~qdrgsm@109.226.79.14] has joined #shogun | 18:16 | |
-!- n4nd0 [~n4nd0@s83-179-44-135.cust.tele2.se] has joined #shogun | 19:09 | |
n4nd0 | hi | 19:10 |
n4nd0 | I am interested in gsoc 2012, I know that the organizations have not even been selected yet | 19:11 |
n4nd0 | but I would like to start getting a bit in shogun in case it is selected I can make a good proposal | 19:11 |
blackburn | n4nd0: hi | 19:13 |
blackburn | n4nd0: well gsoc was not announced too (yet) | 19:13 |
blackburn | :) | 19:13 |
Ram108 | blackburn: hi do u mind if i seek some help? | 19:14 |
n4nd0 | blackburn, yeah, that's right :) | 19:15 |
blackburn | Ram108: sure | 19:15 |
n4nd0 | blackburn, but I guess it will run again this year | 19:15 |
blackburn | n4nd0: yeah, some evidences are here already: http://www.google-melange.com/soc/content/2-0-20120131/images/v2/gsoc/logo/banner-gsoc2012.png | 19:16 |
n4nd0 | blackburn, and anyway, if it doesn't, I would like to learn more about machine learning and this can be a good place to do it :) | 19:16 |
blackburn | n4nd0: how familiar are you with ml? | 19:16 |
Ram108 | i ll catch u tomorrow......... my laptop ll run out of charge in a min and i dont have an outlet to charge | 19:16 |
Ram108 | lol bye have a great day | 19:17 |
n4nd0 | blackburn, I have taken some courses at the university about ml and applied it for the projects in the courses | 19:17 |
blackburn | Ram108: sure, feel free to write emails | 19:17 |
Ram108 | :) | 19:17 |
blackburn | Ram108: see you | 19:17 |
Ram108 | thanks :) | 19:17 |
Ram108 | bye :) | 19:17 |
blackburn | n4nd0: ok so what is the most interesting subarea in ml for you? I'm just curious what to suggest | 19:18 |
-!- Ram108 [~amma@14.96.158.159] has quit [Quit: Ex-Chat] | 19:18 | |
n4nd0 | blackburn, so in one course we talked about kernel functions I found them really cool, I have not applied them to anything though | 19:19 |
n4nd0 | blackburn, I would like to learn more about them | 19:19 |
blackburn | n4nd0: kernels, like in svms? | 19:19 |
n4nd0 | blackburn, yes | 19:20 |
blackburn | hmm we have a lot of them :) | 19:20 |
n4nd0 | blackburn, but are they just for svms or also for more stuff? | 19:20 |
blackburn | n4nd0: well kernels are used in svms very often | 19:20 |
n4nd0 | blackburn, all right, I believed they were also applied to other stuff | 19:21 |
blackburn | they are | 19:21 |
n4nd0 | blackburn, I like svms anyway so it is cool | 19:21 |
blackburn | for example I 'specialize' in dimensionality reduction | 19:21 |
blackburn | and there are some kernel methods | 19:21 |
n4nd0 | ok, I see | 19:22 |
n4nd0 | and apart from the subarea in ml I am interested, what is interesting for the toolbox to have? | 19:22 |
n4nd0 | what are you working on? | 19:23 |
blackburn | n4nd0: I'm author of all dimensionality reduction methods in shogun | 19:23 |
blackburn | an * | 19:23 |
blackburn | :) | 19:23 |
n4nd0 | :) | 19:23 |
blackburn | well this year I'll probably be working on domain adaptation or multitask learning | 19:23 |
blackburn | do you know anything about structured output learning? | 19:24 |
n4nd0 | sincerely not | 19:25 |
blackburn | okk what about langs you are familiar with? | 19:25 |
blackburn | n4nd0: http://shogun-toolbox.org/edrt/ that's the thing I'm mainly working at :) | 19:26 |
n4nd0 | blackburn, I work the most with C++ and matlab/octave | 19:26 |
n4nd0 | blackburn, but I can work in java and python also | 19:26 |
blackburn | I see | 19:26 |
blackburn | well how do you want to start 'shoguning'? :) | 19:27 |
blackburn | you could write some examples or extend with new kernels or distances | 19:27 |
n4nd0 | blackburn, haha that's the big question | 19:27 |
blackburn | or anything similar | 19:28 |
n4nd0 | blackburn, I don't really know, I guess almost anything could work | 19:28 |
n4nd0 | blackburn, that sounds good | 19:28 |
n4nd0 | maybe I can start with examples and later go more into extending some functionality / creating new stuff | 19:29 |
blackburn | sure | 19:29 |
blackburn | with which tasks you was involved? | 19:29 |
blackburn | classification? | 19:29 |
blackburn | in your studies I mean | 19:29 |
blackburn | I would suggest you to reproduce it with shogun | 19:30 |
n4nd0 | I worked on face recognition | 19:30 |
blackburn | oh nice | 19:30 |
blackburn | on which datasets? | 19:30 |
n4nd0 | digit recognition and stuff like that | 19:30 |
blackburn | ok so you know mnist and ORL at least, right? | 19:30 |
n4nd0 | we implemented the paper of Viola and Jones | 19:31 |
n4nd0 | boosting with haar like features | 19:31 |
blackburn | hmm I see | 19:31 |
blackburn | do you have your own code for haar-like features you can integrate? | 19:31 |
n4nd0 | we did the full implementation in matlab | 19:32 |
n4nd0 | so I would have to look for it but yes, I must have that code | 19:32 |
blackburn | ok then my suggestion is: | 19:33 |
blackburn | there could be some script | 19:33 |
blackburn | that 1) downloads some dataset like olivetti or so | 19:33 |
blackburn | 2) constructs these features | 19:33 |
blackburn | 3) classifies with shogun's classifiers | 19:33 |
n4nd0 | compare different classifiers? | 19:35 |
blackburn | well could be nice too | 19:35 |
blackburn | but even with only one classifier it works well too | 19:35 |
n4nd0 | all right | 19:35 |
blackburn | sounds like a plan? | 19:35 |
n4nd0 | sounds like a nice plan :) | 19:35 |
blackburn | you may contact me via blackburn91@gmail.com | 19:36 |
blackburn | or here, in irc | 19:36 |
n4nd0 | ok, thank you | 19:36 |
n4nd0 | :) | 19:37 |
n4nd0 | I will start then refreshing what I did in my lab to compute the features and later decide the classifier to use | 19:37 |
n4nd0 | is it ok to start out with my matlab code? or would it be better in C++ or any other? | 19:42 |
blackburn | n4nd0: does it work with octave? | 19:43 |
blackburn | python is preferrable but no need to rewrite it if it is already working with octave | 19:43 |
n4nd0 | blackburn, I think it should work with octave, I don't think I used anything fancy, but I should try it out first | 19:44 |
n4nd0 | blackburn, so if the part that I have already written to compute the Haar features works well in octave, can it be used from python later? | 19:45 |
blackburn | n4nd0: sure just decompose it to features.m and classify.py :) | 19:46 |
n4nd0 | blackburn, cool | 19:46 |
blackburn | n4nd0: are you from sweden? :) | 19:50 |
n4nd0 | blackburn, no, but I live in Sweden | 19:50 |
n4nd0 | :) | 19:50 |
blackburn | I see | 19:50 |
n4nd0 | blackburn, where are you from? Sweden or Scandinavia maybe? | 19:50 |
blackburn | where are you from then (if it is not a secret)? :) | 19:50 |
n4nd0 | Spain | 19:51 |
blackburn | n4nd0: well my name is Sergey :) | 19:51 |
blackburn | I see | 19:51 |
blackburn | so I'm russian | 19:51 |
n4nd0 | :) | 19:51 |
n4nd0 | blackburn, was it just a good guess :P? or did you find something related to my nickname and Sweden? | 19:52 |
blackburn | from cold country too hah | 19:52 |
n4nd0 | haha yes | 19:52 |
blackburn | n4nd0: Username: ~n4nd0@s83-179-44-135.cust.tele2.se :) | 19:52 |
n4nd0 | haha I see | 19:52 |
-!- f-x [~user@117.216.152.45] has joined #shogun | 19:57 | |
-!- axitkhurana [~akshit@210.212.58.231] has joined #shogun | 19:57 | |
-!- f-x [~user@117.216.152.45] has quit [Remote host closed the connection] | 20:00 | |
-!- axitkhurana [~akshit@210.212.58.231] has quit [Quit: Leaving.] | 20:04 | |
CIA-11 | shogun: Sergey Lisitsyn master * r0be4076 / src/shogun/classifier/ConjugateIndex.cpp : Conjugate index improvement - http://git.io/OdpGXw | 20:26 |
-!- n4nd0 [~n4nd0@s83-179-44-135.cust.tele2.se] has quit [Read error: Operation timed out] | 20:51 | |
-!- n4nd0 [~n4nd0@s83-179-44-135.cust.tele2.se] has joined #shogun | 20:52 | |
n4nd0 | blackburn, so did you have any classifier in particular that could work well with the haar features? | 20:55 |
blackburn | n4nd0: unfortunately I'm not very familiar with haar features | 20:55 |
n4nd0 | blackburn, in the implementation I have I use weak classifiers with boosting | 20:56 |
blackburn | well it is easy to try different ones | 20:56 |
blackburn | with shogun I mean | 20:56 |
n4nd0 | does the toolbox include boosting then? | 20:57 |
blackburn | we only have LPBoost currently | 20:58 |
n4nd0 | I am not familiare with it | 20:58 |
blackburn | me neither | 20:59 |
CIA-11 | shogun: Soeren Sonnenburg master * r013c234 / (32 files in 9 dirs): | 21:02 |
CIA-11 | shogun: Merge pull request #361 from karlnapf/master | 21:02 |
CIA-11 | shogun: KRR/SVR and cross-validation/model-selection - http://git.io/jwTntA | 21:02 |
-!- n4nd0 [~n4nd0@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 272 seconds] | 21:04 | |
-!- n4nd0 [~n4nd0@s83-179-44-135.cust.tele2.se] has joined #shogun | 21:16 | |
CIA-11 | shogun: Sergey Lisitsyn master * re251619 / (2 files): Added confusion matrix routine - http://git.io/ZhEhSQ | 22:01 |
-!- blackburn [~qdrgsm@109.226.79.14] has quit [Quit: Leaving.] | 22:20 | |
shogun-buildbot | build #122 of nightly_default is complete: Failure [failed test_1] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/122 | 22:31 |
-!- n4nd0 [~n4nd0@s83-179-44-135.cust.tele2.se] has quit [Ping timeout: 240 seconds] | 23:59 | |
--- Log closed Thu Feb 02 00:00:19 2012 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!