--- Log opened Wed Apr 25 00:00:19 2012 | ||
-!- Marty28 [~marty@158.181.78.199] has quit [Quit: Colloquy for iPad - http://colloquy.mobi] | 00:05 | |
-!- n4nd0 [5aac282c@gateway/web/freenode/ip.90.172.40.44] has joined #shogun | 00:12 | |
n4nd0 | good evening guys | 00:12 |
---|---|---|
n4nd0 | from Spain | 00:12 |
-!- harshit_ [~harshit@182.68.43.52] has joined #shogun | 00:16 | |
n4nd0 | I will see you tom around here, good night! | 00:23 |
-!- n4nd0 [5aac282c@gateway/web/freenode/ip.90.172.40.44] has quit [Quit: Page closed] | 00:23 | |
-!- gsomix [~gsomix@188.168.5.178] has quit [Read error: Operation timed out] | 00:27 | |
-!- harshit_ [~harshit@182.68.43.52] has quit [Quit: Leaving] | 01:01 | |
-!- vikram360 [~vikram360@117.192.161.188] has quit [Ping timeout: 244 seconds] | 03:04 | |
-!- pluskid [~pluskid@li381-56.members.linode.com] has joined #shogun | 04:42 | |
* pluskid bought more memory for the laptop, need to install a 64bit system now | 04:43 | |
-!- pluskid [~pluskid@li381-56.members.linode.com] has quit [Client Quit] | 04:47 | |
-!- pluskid [~chatzilla@111.120.57.192] has joined #shogun | 05:23 | |
-!- vikram360 [~vikram360@117.192.161.188] has joined #shogun | 05:58 | |
-!- pluskid [~chatzilla@111.120.57.192] has quit [Ping timeout: 252 seconds] | 06:04 | |
-!- vikram360 [~vikram360@117.192.161.188] has quit [Ping timeout: 260 seconds] | 08:01 | |
-!- uricamic [9320543b@gateway/web/freenode/ip.147.32.84.59] has joined #shogun | 08:49 | |
-!- pluskid [~chatzilla@111.120.57.192] has joined #shogun | 08:52 | |
-!- blackburn [5bdfb203@gateway/web/freenode/ip.91.223.178.3] has joined #shogun | 08:53 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 09:06 | |
-!- n4nd0 [d480fcac@gateway/web/freenode/ip.212.128.252.172] has joined #shogun | 09:08 | |
n4nd0 | good morning! | 09:08 |
blackburn | n4nd0: hey how was your flight? | 09:11 |
n4nd0 | blackburn: it went good | 09:11 |
pluskid | Is it Apr. 25 or Apr. 26 now in your time zone? | 09:11 |
n4nd0 | blackburn: first flight ever seen with so many Smith charts :P | 09:11 |
blackburn | pluskid: 25 | 09:12 |
n4nd0 | 25th | 09:12 |
pluskid | so the kick off meeting is on 26th, right? | 09:12 |
blackburn | pluskid: 27 for you probably | 09:12 |
blackburn | orr | 09:12 |
blackburn | no | 09:12 |
blackburn | are you in utc+8? | 09:12 |
pluskid | yes | 09:13 |
n4nd0 | sonney2k: hey! about the kickoff meeting, my exam starts at 13:00 UTC tomorrow, I hope there is no trouble with that since it starts 11:00 UTC | 09:13 |
n4nd0 | sonney2k: do you happen to know how long will it take? | 09:13 |
blackburn | n4nd0: I don't think it would take more than half of hour | 09:13 |
pluskid | n4nd0: good luck with your exam | 09:13 |
-!- sonne|work [~sonnenbu@194.78.35.195] has joined #shogun | 09:13 | |
blackburn | pluskid: wait but is it 26th in your timezone? | 09:13 |
n4nd0 | pluskid: thank you! | 09:14 |
n4nd0 | blackburn: great | 09:14 |
blackburn | I don't think so :) | 09:14 |
blackburn | it should be 15 now? | 09:14 |
pluskid | 25th, now, in my time zone | 09:14 |
n4nd0 | I will ask sonne|work also just in case | 09:14 |
pluskid | afternoon of 25th | 09:14 |
sonne|work | pluskid: looks like you didn't read the subject of the email | 09:14 |
sonne|work | pluskid: it said 26th | 09:14 |
blackburn | sonne|work: no he is confused with timezones | 09:14 |
blackburn | so we check watches :D | 09:15 |
pluskid | sonne|work: oops, you are right | 09:15 |
-!- vikram360 [~vikram360@117.192.161.188] has joined #shogun | 09:15 | |
pluskid | thanks! | 09:15 |
sonne|work | we are all on the same day (25th now) | 09:15 |
n4nd0 | sonne|work: is it a problem if my exam tomorrow starts two hours after the kickoff is scheduled? I mean in terms of how long will it be | 09:15 |
blackburn | not two hours definitely | 09:16 |
sonne|work | it is clear from the beginning that not everyone can attend / some will have to leave early | 09:19 |
sonne|work | but I have to leave after 1 hr myself so... | 09:19 |
n4nd0 | no worries then :) | 09:20 |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking] | 09:27 | |
-!- pluskid [~chatzilla@111.120.57.192] has quit [Remote host closed the connection] | 09:30 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 09:47 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 09:47 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 09:47 | |
-!- n4nd0 [d480fcac@gateway/web/freenode/ip.212.128.252.172] has quit [Ping timeout: 245 seconds] | 09:59 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking] | 09:59 | |
blackburn | sonne|work: just a note - we need to discuss cases when one is stucked and unhappy | 10:08 |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 10:08 | |
blackburn | during kickoff I think | 10:08 |
-!- n4nd0 [~n4nd0@wifi-cct-cisco-r252-172.cct.urjc.es] has joined #shogun | 10:08 | |
sonne|work | blackburn: make a list and remind me if I forget sth :) | 10:10 |
blackburn | sonne|work: shogun meeting @ berlin, middle of august? | 10:13 |
n4nd0 | +1000 to that suggestion :) | 10:13 |
sonne|work | heh I am afraid not this year | 10:14 |
blackburn | why? | 10:14 |
blackburn | http://nuit-blanche.blogspot.com/2012/04/compressive-sensing-this-week.html | 10:23 |
blackburn | sonne|work: n4nd0: ^ | 10:23 |
blackburn | it is pretty unusual to read about myself hehe | 10:23 |
n4nd0 | cool | 10:24 |
sonne|work | blackburn: hehe :) | 10:24 |
n4nd0 | who is the blogger? | 10:24 |
blackburn | no idea | 10:25 |
n4nd0 | haha you're famous | 10:26 |
-!- karlnapf [~heiko@host86-180-44-86.range86-180.btcentralplus.com] has joined #shogun | 10:42 | |
-!- emre-away is now known as emrecelikten | 10:46 | |
-!- karlnapf [~heiko@host86-180-44-86.range86-180.btcentralplus.com] has left #shogun [] | 10:50 | |
blackburn | sonne|work: don't you think tutorial should be a submodule? | 10:52 |
blackburn | like data | 10:52 |
sonne|work | blackburn: of course but you did fail putting your tutorial there :D | 10:54 |
blackburn | sonne|work: tutorial of non-existent toolkit? should take more time to finish it | 10:54 |
sonne|work | yes sure - we have to start somehow right? | 10:56 |
blackburn | sonne|work: what about cherry-picking some stuff from your thesis? | 11:00 |
sonne|work | makes sense but still lots and lots of work... | 11:01 |
blackburn | sonne|work: can I obtain source? | 11:02 |
sonne|work | sure I can give it to you | 11:02 |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Read error: Connection reset by peer] | 11:02 | |
blackburn | sonne|work: you thesis is well-written so it makes sense to cherry-pick some parts like string kernels | 11:03 |
-!- wiking [~wiking@we02c096.ugent.be] has joined #shogun | 11:03 | |
-!- wiking [~wiking@we02c096.ugent.be] has quit [Changing host] | 11:03 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 11:03 | |
blackburn | I should obtain chris' thesis heh | 11:04 |
wiking | damn my machine crashed... :( | 11:07 |
blackburn | wiking: do you think dot product of homogay kernel map can be optimized somehow? | 11:08 |
blackburn | sh I can't stop call it homogay it is not funny at all :D | 11:08 |
wiking | blackburn: well homo gay kernel map should be first extended to the new stuff ;) | 11:08 |
wiking | blackburn: remember the 2 papers i've sent | 11:08 |
blackburn | wiking: I mean they use sin and cos right? | 11:09 |
wiking | yes | 11:09 |
blackburn | so in dot product we could reduce it probably? | 11:09 |
wiking | ? | 11:09 |
blackburn | sin(x)*sin(y) = 2 sin(x+y) or so? | 11:09 |
blackburn | 1/2 | 11:09 |
blackburn | heh | 11:09 |
blackburn | no idea if it worths trying | 11:10 |
wiking | mmmm | 11:11 |
wiking | now i'm starting to see what you mean ;) | 11:11 |
wiking | could be interesting | 11:11 |
blackburn | wiking: actually it seems we will use some identical multipliers | 11:12 |
blackburn | so may be we can just precompute squares? | 11:12 |
wiking | eh? | 11:14 |
wiking | ok now i'm lost blackburn | 11:14 |
wiking | so expand your story :) | 11:14 |
blackburn | each feature is being multiplied | 11:14 |
blackburn | with sin and cos and 1? | 11:14 |
wiking | not exactly this straightforward | 11:15 |
blackburn | ah I checked code and it is not as simple as I thought hehe | 11:16 |
blackburn | wiking: okay I am lost :D | 11:17 |
blackburn | but the idea stays - is dot product easier to compute than full mapping? | 11:17 |
wiking | it needs to be checked | 11:18 |
blackburn | wiking: have you time for that? | 11:18 |
wiking | yeah no worries | 11:19 |
wiking | i can hope on that train ;P | 11:19 |
blackburn | I'm going to start with slep porting tomorrow | 11:19 |
wiking | yeey | 11:20 |
CIA-64 | shogun: Heiko Strathmann master * rb47d55b / src/shogun/features/SimpleFeatures.cpp : fixed chache bug reported by - http://git.io/su1RUQ | 11:27 |
CIA-64 | shogun: Heiko Strathmann master * r0afeee4 / src/shogun/features/SimpleFeatures.cpp : | 11:27 |
CIA-64 | shogun: Merge pull request #485 from karlnapf/master | 11:27 |
CIA-64 | shogun: bug fix reported by Koen van de Sande - http://git.io/nY3K5A | 11:27 |
emrecelikten | Hey guys, what's the primary difference between a Markov chain and a Markov random field? One is a probabilistic FSM while the other is like a undirected Bayesian network? | 11:27 |
n4nd0 | emrecelikten, are you just wondering or do you have to choose between both of them for something in particular? | 11:31 |
emrecelikten | n4nd0: Wondering. Could MRFs be applied to speech/music domain like HMMs for example | 11:32 |
emrecelikten | I think so | 11:32 |
n4nd0 | emrecelikten, I am not sure, I have never really worked with MRFs far beyond definition and toy examples | 11:33 |
blackburn | emrecelikten: markov chain is a chain right? :) | 11:33 |
blackburn | while random field is a graph structure | 11:33 |
n4nd0 | emrecelikten, I don't really see direct application of MRF to it but it might be | 11:33 |
n4nd0 | blackburn, isn't it a chain a particular case of directed weighted graph? | 11:34 |
blackburn | n4nd0: yes but it is a chain in markov chains - not a graph | 11:34 |
emrecelikten | Hmm thanks | 11:35 |
n4nd0 | blackburn, I partially agree, I think that a graph is something more general than a chain | 11:36 |
blackburn | yes exactly what I mean | 11:36 |
n4nd0 | but I'd say that a chain is a graph | 11:36 |
blackburn | it is | 11:36 |
blackburn | acyclic with nodes with only two or one output edge ? | 11:37 |
blackburn | something like that I guess | 11:38 |
n4nd0 | a Markov chain can be cyclic I think | 11:39 |
blackburn | right sorry | 11:39 |
n4nd0 | no problem :) | 11:39 |
blackburn | but no multiple outputs | 11:39 |
n4nd0 | yeah, they can have multiple | 11:39 |
n4nd0 | if today is sunny | 11:40 |
blackburn | wait but when I am fucked up - what is the difference then? | 11:40 |
n4nd0 | tomorrow might be rainy, sunny, cloudy or pouring cats and dogs for example | 11:40 |
n4nd0 | there you have more than two output edges | 11:40 |
n4nd0 | the difference between which ones? | 11:40 |
blackburn | MC and MRF | 11:41 |
n4nd0 | I think MRF are something similar to Bayesian nets | 11:41 |
n4nd0 | so for example the graph that represent them is not weighted | 11:41 |
n4nd0 | there must be more differences | 11:42 |
n4nd0 | at least I have found them in different realms | 11:42 |
blackburn | damn I should get back to markov stuff and study it :D | 11:42 |
blackburn | emrecelikten: probably I am wrong | 11:42 |
n4nd0 | I think MRF is more advanced stuff | 11:44 |
n4nd0 | one type of probabilistic graphical model | 11:44 |
emrecelikten | blackburn: It's fine :) | 11:44 |
n4nd0 | I have to learn more of them for the project since I'll have to implement conditional random fields | 11:44 |
emrecelikten | blackburn: Longtimenostudy effect :P :D | 11:45 |
blackburn | emrecelikten: oh it's been a long without some effects and paradoxes ;) | 11:45 |
emrecelikten | MRF supports higher dimensions I think | 11:45 |
emrecelikten | blackburn: :P | 11:45 |
-!- vikram360 [~vikram360@117.192.161.188] has quit [Read error: Connection reset by peer] | 11:46 | |
-!- vikram360 [~vikram360@117.192.161.188] has joined #shogun | 11:46 | |
* blackburn is reading bishop book | 11:48 | |
emrecelikten | n4nd0: Are you taking PGM class from Coursera? | 11:49 |
n4nd0 | emrecelikten, I didn't get time to do it now but I would like to follow it during summer | 11:49 |
n4nd0 | emrecelikten, are you? | 11:49 |
emrecelikten | n4nd0: Yes, but its difficult | 11:50 |
n4nd0 | emrecelikten, do you recommend it? | 11:50 |
-!- uricamic [9320543b@gateway/web/freenode/ip.147.32.84.59] has quit [Ping timeout: 245 seconds] | 11:50 | |
emrecelikten | n4nd0: Just failed my first programming assignment last night, very time consuming (and slightly annoying)... But its a good class | 11:51 |
n4nd0 | oh | 11:51 |
emrecelikten | n4nd0: I have seen a "proof by example" though, so I should warn you in advance if you like rigor :D | 11:54 |
-!- puffin444 [230bf329@gateway/web/freenode/ip.35.11.243.41] has joined #shogun | 12:00 | |
-!- puffin444 [230bf329@gateway/web/freenode/ip.35.11.243.41] has left #shogun [] | 12:00 | |
-!- gsomix [~gsomix@85.26.164.208] has joined #shogun | 12:24 | |
gsomix | sonney2k, around? | 12:29 |
gsomix | sonne|work, | 12:29 |
-!- vikram360 [~vikram360@117.192.161.188] has quit [Ping timeout: 260 seconds] | 12:30 | |
n4nd0 | gtg guys, see you later! | 12:31 |
blackburn | n4nd0: see you | 12:31 |
-!- n4nd0 [~n4nd0@wifi-cct-cisco-r252-172.cct.urjc.es] has quit [Quit: Ex-Chat] | 12:33 | |
-!- vikram360 [~vikram360@117.192.163.57] has joined #shogun | 12:47 | |
-!- gsomix [~gsomix@85.26.164.208] has quit [Ping timeout: 272 seconds] | 12:51 | |
-!- blackburn [5bdfb203@gateway/web/freenode/ip.91.223.178.3] has quit [Quit: Page closed] | 13:08 | |
-!- gsomix [~gsomix@188.168.5.195] has joined #shogun | 13:14 | |
sonne|work | gsomix: yes? | 13:23 |
gsomix | sonne|work, I'm so sorry, but I not able go to kickoff meeting, because I have classes at this time. What to do? I can prepare text for tomorrow, if it's needed. | 13:24 |
gsomix | sonne|work, moin | 13:24 |
sonne|work | gsomix: not a problem - you can read the chat logs later | 13:24 |
sonne|work | gsomix: it was clear from the beginning that not everyone can attend | 13:24 |
gsomix | sonne|work, thanks a lot. :) | 13:24 |
sonne|work | it is very hard to schedule a meeting for 16 participants w/o planning years into the future :) | 13:25 |
gsomix | sonne|work, huh, I see. | 13:26 |
-!- harshit_ [~harshit@182.68.43.52] has joined #shogun | 13:39 | |
* gsomix http://www.planetaryresources.com/mission/ | 13:48 | |
-!- vikram360 [~vikram360@117.192.163.57] has quit [Ping timeout: 240 seconds] | 13:49 | |
-!- vikram360 [~vikram360@117.192.189.186] has joined #shogun | 13:49 | |
emrecelikten | gsomix: So it was truly asteroid mining then | 13:51 |
emrecelikten | Why don't we start constructing a Dyson sphere instead :P | 13:52 |
-!- harshit_ [~harshit@182.68.43.52] has quit [Remote host closed the connection] | 14:05 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking] | 14:05 | |
-!- pluskid_ [~chatzilla@173.254.214.60] has joined #shogun | 14:16 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 14:17 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 14:17 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 14:17 | |
-!- pluskid_ is now known as pluskid | 14:19 | |
wiking | mmm | 14:24 |
wiking | pluskid: here? | 14:24 |
pluskid | wiking: hi | 14:25 |
wiking | oh great hi! | 14:26 |
wiking | so i was just wondering about your refactoring of multi class | 14:26 |
wiking | as i'm now implementing the latent svm | 14:27 |
pluskid | so what's your question? | 14:27 |
wiking | it's a derivative of LinearMachine | 14:27 |
wiking | and as a first step it's binary | 14:27 |
pluskid | hmm | 14:28 |
wiking | but if i want to extend it to multi class machine, do we have line LinearMulticlass? | 14:28 |
pluskid | does your latent svm support multiclass *naturally* ? | 14:28 |
pluskid | I mean | 14:29 |
pluskid | for example, traditional svm only support binary case | 14:29 |
pluskid | but we can use one-vs-rest or one-vs-one to make it multiclass | 14:29 |
-!- blackburn [~qdrgsm@188.168.2.179] has joined #shogun | 14:29 | |
wiking | pluskid: no it does not support multi class naturally | 14:29 |
-!- PhilTillet [~Philippe@vir78-1-82-232-38-145.fbx.proxad.net] has joined #shogun | 14:29 | |
pluskid | if your latent svm also rely one one-vs-res or one-vs-one for multiclass | 14:30 |
wiking | but i'd like it to be able to use for multi class case as well | 14:30 |
pluskid | then I guess you don't have to do anything | 14:30 |
wiking | ah ok | 14:30 |
pluskid | because MulticlassLinearMachine can accept a "base linear machine" | 14:30 |
wiking | yeah as a first step i guess it should be OvR OvO | 14:30 |
pluskid | and turn that base linear machine into multiclass | 14:30 |
wiking | ok | 14:30 |
wiking | that's great | 14:31 |
pluskid | there's an example code I guess | 14:31 |
wiking | and in case i can extend it to a true multi class machine? | 14:31 |
wiking | i.e. like a crammer-singer | 14:31 |
pluskid | if it becomes a true multiclass machine, you will derive a subclass from MulticlassLinearMachine | 14:31 |
pluskid | it cannot be subclass of both linearmachine and multiclasslinearmachine | 14:32 |
pluskid | that's our problem | 14:32 |
pluskid | also happens in MulticlassSVM | 14:32 |
wiking | aha | 14:32 |
pluskid | cannot be subclass of SVM and MultickassKernelmachine simultaneously | 14:32 |
wiking | but why not? :) | 14:32 |
pluskid | we do not like multiple inheritance here | 14:32 |
wiking | i mean i guess the current design does not allow it | 14:32 |
wiking | why? :) | 14:32 |
-!- n4nd0 [~n4nd0@wifi-cct-cisco-r252-4.cct.urjc.es] has joined #shogun | 14:33 | |
wiking | it's not that of a 'big magic' :) | 14:33 |
pluskid | because sonney2k hate multiple inheritance :p | 14:33 |
pluskid | joking haha | 14:33 |
pluskid | I guess if we really want multiple inheritance, we will need "virtual inheritance" | 14:34 |
wiking | :>>> | 14:34 |
wiking | not necesarrily | 14:34 |
wiking | but of course it depends a lot on the base classes | 14:34 |
pluskid | because both super class has a common super class (CSGObject) | 14:34 |
wiking | ah yeah | 14:34 |
wiking | heheh in that case yes | 14:35 |
pluskid | the inheritance graph becoms complicated | 14:35 |
wiking | yeah a diamond inharitance | 14:35 |
pluskid | yes, like std::iostream ? :p | 14:36 |
wiking | anyhow that's another story for later | 14:36 |
wiking | first i want to see how it performs with a OvR setup | 14:36 |
wiking | and as you said that should be fairly easy to do | 14:37 |
pluskid | wiking: check this example | 14:37 |
pluskid | https://github.com/shogun-toolbox/shogun/blob/master/examples/undocumented/libshogun/classifier_multiclasslinearmachine.cpp | 14:37 |
pluskid | it use a LibLinear as a base classifier | 14:37 |
pluskid | and use OneVsOne to turn it into multiclass | 14:37 |
pluskid | what you need is to new a CMulticlassOneVsOneStrategy() haha | 14:37 |
pluskid | there will be ECOC support soon :) | 14:41 |
wiking | great! | 14:41 |
wiking | thanks a lot i'll do the testing with this | 14:41 |
wiking | i mean based on this | 14:41 |
pluskid | :) | 14:41 |
wiking | yeeey this should be fun to see how it performs :))) | 14:41 |
pluskid | I didn't tested the performance of the shogun machines really | 14:41 |
pluskid | be back in a while | 14:41 |
blackburn | wiking: hey what is your issue? | 14:44 |
wiking | blackburn: well i just want to have multi class support for latent svm ;) | 14:45 |
blackburn | wiking: hmm can you describe in a few words how does it extend? | 14:46 |
wiking | well currently as you can see only with OvR | 14:46 |
wiking | but i must check whether i can use crammer-singer | 14:47 |
wiking | but as for now i don't see a problem with it | 14:47 |
blackburn | wiking: crammer-singer is done in both ocas and liblinear multiclass svms | 14:47 |
wiking | yeah i know | 14:47 |
wiking | and since i currently use ocas | 14:47 |
wiking | it should be fine | 14:48 |
blackburn | wiking: I already asked vojtech about some improvement of ocas btw | 14:48 |
wiking | but then i guess i have to the the 'reimplementation' of latent svm | 14:48 |
blackburn | I need to solve 0.5||w||^2 + <w0,w> + C \sum_i \xi_i problems | 14:48 |
wiking | as current implementation is derived from linearmachine | 14:49 |
blackburn | (domain adaptation) | 14:49 |
blackburn | wiking: yes sad truth | 14:49 |
wiking | ah w_0 is the input learned w? | 14:49 |
blackburn | linear machine is pretty 'far away' from linearmulticlassmachine | 14:49 |
blackburn | wiking: yes kind of pre-svm | 14:49 |
blackburn | source domain svm | 14:49 |
wiking | yep | 14:49 |
blackburn | (but in multiclass case) | 14:49 |
blackburn | so \sum_m w_m everywhere | 14:50 |
wiking | btw: i see that we have SVMOcas for binary case | 14:52 |
wiking | but there's no such for multi class case | 14:52 |
wiking | ah no | 14:53 |
wiking | i'm wrong | 14:53 |
wiking | we do | 14:53 |
blackburn | wiking: multiclass/MulticlassOCAS | 14:53 |
wiking | yes yes found it sorry | 14:53 |
blackburn | wiking: I have pretty recently implemented it actually | 14:53 |
wiking | that should be interesting to test | 14:53 |
wiking | i'll have to change the api of my implementation as far as i can see | 14:54 |
wiking | since maybe the best thing is to somehow supply the svm you want to use | 14:54 |
wiking | for latent svm | 14:54 |
wiking | but i'm wondering how will an s-svm look like | 14:54 |
wiking | n4nd0 will tell me more about that soon ;) | 14:54 |
blackburn | yeah once he is finished with some univ stuff :) | 14:55 |
n4nd0 | yeah :) | 14:55 |
wiking | :>>> | 14:55 |
n4nd0 | wiking, in any case, yesterday I devoted some time to read some of the Struct SVM's doc | 14:55 |
-!- pluskid_ [~chatzilla@111.120.57.192] has joined #shogun | 14:55 | |
n4nd0 | wiking, I like some of the diagrams they have there and I think that it is exactly what we want to do in shogun | 14:56 |
wiking | :)))) | 14:56 |
n4nd0 | I am waiting for an answer from Nico to know that though | 14:56 |
n4nd0 | wiking, have you taken a look to Struct SVM? | 14:56 |
wiking | yep | 14:56 |
n4nd0 | cool | 14:56 |
-!- pluskid [~chatzilla@173.254.214.60] has quit [Read error: Connection reset by peer] | 14:56 | |
wiking | currently i'm doing runs | 14:56 |
blackburn | n4nd0: which diagrams? | 14:57 |
wiking | my implementation vs the other latent implementation | 14:57 |
-!- pluskid_ is now known as pluskid | 14:57 | |
n4nd0 | blackburn, a moment | 14:57 |
wiking | that is based on structsvm | 14:57 |
wiking | but i just have -1,1 labeling so i don't use SO | 14:57 |
n4nd0 | blackburn, just scroll down a bit http://svmlight.joachims.org/svm_struct.html | 14:57 |
n4nd0 | wiking, I have to think about how to fit into shogun the labels hierarchy we need for SO | 14:58 |
blackburn | hmm where? | 14:58 |
wiking | n4nd0: i'm changing as well labels... | 14:59 |
wiking | since i have latent output | 14:59 |
wiking | so it might be good to talk about it when u have some time | 14:59 |
blackburn | n4nd0: are you sure it is the link? | 14:59 |
n4nd0 | blackburn, maybe diagrams is a bad word to refer to them | 14:59 |
blackburn | I see only text :) | 15:00 |
n4nd0 | blackburn, the "coloured boxes" on the left with arrows in between | 15:00 |
n4nd0 | a wait | 15:00 |
n4nd0 | http://www.cs.cornell.edu/~tomf/svmpython2/ | 15:00 |
blackburn | ah | 15:01 |
n4nd0 | the link remains the same in the browser's textbox for the address :D | 15:01 |
blackburn | yes any kind of diagrams help | 15:01 |
n4nd0 | I think that SVM struct is a good reference for what we want to do | 15:02 |
blackburn | n4nd0: yes yes | 15:02 |
n4nd0 | but I wonder if I got something wrong since it was not referenced in the project ideas | 15:02 |
blackburn | pretty general | 15:02 |
n4nd0 | that's why I asked Nico about it, just to be sure | 15:02 |
blackburn | hmm I don't think thorsten has fucked up | 15:03 |
blackburn | :D | 15:03 |
n4nd0 | :D | 15:03 |
blackburn | n4nd0: wiking: it should be of interest right? http://svmlight.joachims.org/svm_struct.html | 15:05 |
wiking | blackburn: well yeah that's a typical structural svm | 15:05 |
blackburn | I am very excited about having all that jazz in shogun | 15:06 |
wiking | hehehe | 15:06 |
n4nd0 | it's going to be very cool | 15:07 |
pluskid | what will you use as labels in SO learning? | 15:10 |
pluskid | I mean the current CLabels wouldn't be enough to describe structures | 15:10 |
wiking | pluskid: yeah it needs to be extended | 15:10 |
wiking | btw when are we planning to do a release? | 15:10 |
wiking | i mean shogun release | 15:10 |
pluskid | any concrete idea? I'm thinking of extending CLabels for multiclass, too. though it can support multiclass in the current status | 15:11 |
blackburn | pluskid: yes would be nice to have multiclass lables | 15:11 |
blackburn | probabilities is the thing we can store there | 15:11 |
blackburn | wiking: that would be hard to release before end of gsoc currently | 15:12 |
wiking | heheh ok | 15:12 |
wiking | afaik the last release was 8 months ago | 15:12 |
blackburn | however this can be done | 15:12 |
wiking | and there's a lot of improvement/fixes since then | 15:12 |
blackburn | yes but we are in unstable state | 15:12 |
wiking | true | 15:12 |
blackburn | all is broken | 15:12 |
blackburn | :D | 15:12 |
wiking | api is changing a lot :)) | 15:12 |
pluskid | have you considered some benchmark test for our machines in shogun? | 15:13 |
blackburn | I'd call it shogun 2.0 | 15:13 |
blackburn | :D | 15:13 |
blackburn | pluskid: what to benchmark? | 15:13 |
pluskid | for examples, all the examples only run classification etc. | 15:13 |
wiking | pluskid: i do with my own... | 15:13 |
pluskid | but do not evaluate | 15:13 |
blackburn | performance evaluation? | 15:13 |
pluskid | yes | 15:13 |
pluskid | that would also be a good reference when we are doing refactoring of the code | 15:14 |
pluskid | if the performance are different before and after, there must be something wrong | 15:14 |
blackburn | ah you mean to compare things using there benchmarks? | 15:14 |
pluskid | yes | 15:14 |
blackburn | I don't think it would work very well | 15:15 |
pluskid | and this is also useful for the users to choose machines to use | 15:15 |
pluskid | blackburn: why? | 15:15 |
blackburn | pluskid: what would you measure for classifiers? | 15:15 |
pluskid | for example, accuracy on a standard dataset | 15:15 |
blackburn | but why not to compare learned parameters? | 15:16 |
pluskid | for the purpose of refactoring, your way is better. But performance is more intuitive to see, and has other uses | 15:17 |
pluskid | I mean, for example, I'm a user | 15:18 |
pluskid | of shogun | 15:18 |
blackburn | pluskid: where would you like to add it? | 15:18 |
pluskid | I downloaded the shogun, compile it and run the example code | 15:18 |
pluskid | when I see it produces performance as indicated in the reference, I would be sure that everything is fine, and I can start my work | 15:18 |
pluskid | blackburn: no idea yet, just thinking about this :p | 15:19 |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking] | 15:23 | |
pluskid | something like this: http://yuml.me/f14284c4.jpg | 15:29 |
blackburn | pluskid: hehehe | 15:29 |
blackburn | pluskid: binary=real | 15:30 |
blackburn | ah | 15:30 |
blackburn | regression | 15:30 |
pluskid | hmm | 15:30 |
blackburn | sorry no | 15:30 |
blackburn | makes sense | 15:30 |
blackburn | pluskid: btw multiple output regression is something real too | 15:30 |
blackburn | so RealVectorLabels | 15:30 |
-!- wiking [~wiking@we02c096.ugent.be] has joined #shogun | 15:30 | |
-!- wiking [~wiking@we02c096.ugent.be] has quit [Changing host] | 15:30 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 15:30 | |
pluskid | hmm, make sense | 15:31 |
blackburn | the problem is | 15:31 |
blackburn | how to implement it | 15:31 |
blackburn | we have get_labels() | 15:31 |
blackburn | what should it return? | 15:31 |
blackburn | I see templates as a choice actually | 15:31 |
pluskid | get_labels() is a function of what object? | 15:31 |
blackburn | of labels | 15:31 |
blackburn | or get_label(i) | 15:31 |
pluskid | sonney2k might hate template? :p | 15:31 |
blackburn | no, I don't think | 15:32 |
blackburn | I can't see any other way | 15:32 |
blackburn | pluskid: how did you draw it? | 15:32 |
pluskid | blackburn: http://yuml.me/edit/f14284c4 | 15:33 |
blackburn | all the uml stuff is so damn heavy | 15:33 |
blackburn | that's fscking awesome | 15:33 |
pluskid | haha | 15:33 |
pluskid | this is not *drawing* , though :p | 15:34 |
blackburn | yeah | 15:34 |
wiking | hehehe | 15:35 |
wiking | you can include LatentLabels as well ;P | 15:36 |
wiking | and it's not drawing for me | 15:36 |
pluskid | wiking: what would a latent label looks like? | 15:36 |
wiking | pluskid: well it can be basically anything | 15:36 |
wiking | it's user defined | 15:36 |
pluskid | hmm | 15:37 |
wiking | since the latent variable should be defined by the user | 15:37 |
blackburn | we are going to have great hierarchy there | 15:37 |
blackburn | I am afraid we need shogun/labels | 15:37 |
pluskid | yes, I feel very strange the first time when I found CLabels in features folder :D | 15:37 |
blackburn | sonney2k: it is going to be tough to support back compatibility for this release | 15:37 |
blackburn | everything will be broken | 15:38 |
wiking | pluskid: did the yuml.me code generate you the diagram? | 15:38 |
pluskid | maybe everything is already broken ? :p | 15:38 |
blackburn | is already and will be more | 15:38 |
pluskid | wiking: yes, scroll down the page you will see links to the pictures | 15:39 |
pluskid | of various type, png, pdf, jpg, svg, etc | 15:39 |
wiking | lool | 15:39 |
wiking | yeah that works | 15:39 |
wiking | but not within the page | 15:39 |
blackburn | pluskid: would you mind describing multiclass part @ wikipedia? | 15:46 |
pluskid | wikipedia of shogun? | 15:46 |
blackburn | yes | 15:46 |
blackburn | http://en.wikipedia.org/wiki/Shogun_(toolbox) | 15:46 |
wiking | http://en.wikipedia.org/wiki/Shogun_(toolbox) | 15:47 |
wiking | mmm | 15:47 |
wiking | funny | 15:47 |
wiking | it missed for me the ) | 15:47 |
pluskid | hmm, OK | 15:47 |
pluskid | I'll update it after I add each new multiclass support | 15:48 |
pluskid | algorithm | 15:48 |
-!- vikram360 [~vikram360@117.192.189.186] has quit [Ping timeout: 260 seconds] | 15:49 | |
blackburn | I'm going to add some dim reduction part there once | 15:49 |
blackburn | simply lets make it best article of wikipedia | 15:49 |
pluskid | haha | 15:49 |
pluskid | ambitious goal | 15:49 |
blackburn | no - we know this field | 15:50 |
blackburn | :) | 15:50 |
blackburn | n4nd0: wiking: join me in this effort ;) | 15:52 |
wiking | hehehehe | 15:52 |
pluskid | here's some pseudo code: https://gist.github.com/2489859 | 15:52 |
pluskid | of new CLabels | 15:52 |
blackburn | we need promotion - it worths doing | 15:52 |
pluskid | but the problem is that | 15:52 |
blackburn | in the end your names will be well known :) | 15:53 |
pluskid | if we use template here, they no longer share a *single* super class | 15:53 |
blackburn | pluskid: I would skip binary labels | 15:53 |
blackburn | it is inconvient at some point | 15:54 |
blackburn | i.e. you train svm using binary | 15:54 |
blackburn | but it will always return real | 15:54 |
pluskid | hmm | 15:54 |
pluskid | that's why we have both LabelStoreType and LabelType template variable | 15:55 |
blackburn | ah | 15:55 |
blackburn | brb | 15:55 |
-!- harshit_ [~harshit@182.68.43.52] has joined #shogun | 16:15 | |
blackburn | wiking: last release was on 1, December 2011 btw | 16:37 |
wiking | :> | 16:39 |
-!- emrecelikten [~emre@176.40.232.233] has quit [Quit: Leaving.] | 16:47 | |
-!- harshit_ [~harshit@182.68.43.52] has quit [Quit: Leaving] | 17:07 | |
n4nd0 | blackburn, wiking: Nico has approved the diagrams I linked you guys, svm struct may be an inspiration for our SO base | 17:14 |
wiking | hehehe | 17:14 |
blackburn | n4nd0: heh | 17:14 |
blackburn | nice | 17:14 |
wiking | can u do a sketch of that | 17:14 |
wiking | how that actually would look like in case of shogun | 17:14 |
* wiking is lazy ;) | 17:14 | |
n4nd0 | wiking, yes, we have discussed and would like to prepare some UML diagrams and stuff | 17:15 |
wiking | i mean in my case | 17:15 |
wiking | i'm using psi function as well | 17:15 |
wiking | but it's | 17:15 |
wiking | PSI(x,y,h) | 17:15 |
n4nd0 | what is your h? | 17:15 |
wiking | so "maybe" it's a good idea to sync | 17:16 |
wiking | h = latent variable | 17:16 |
n4nd0 | yes, it's a good idea | 17:16 |
n4nd0 | I don't know much about latent variables but | 17:17 |
wiking | n4nd0: psi is defined by the user right? | 17:17 |
wiking | n4nd0: i mean in case of SO | 17:17 |
n4nd0 | wiking, yes | 17:17 |
wiking | so you'll have somewhere in the class of features or something | 17:17 |
wiking | that is user defined | 17:17 |
n4nd0 | wiking, just a kind of interface hast to be followed | 17:17 |
wiking | currently my CLatentFeatures is something like that | 17:17 |
wiking | you add x, y and h | 17:17 |
wiking | and you have to define PSI | 17:17 |
wiking | so maybe that could be used | 17:18 |
wiking | just then we need to rename it | 17:18 |
wiking | so that it works for SO | 17:18 |
wiking | PSI(x,y) | 17:18 |
n4nd0 | I think that for SO h should be within x | 17:19 |
n4nd0 | I mean, the general algorithm shouldn't really case whether some features there are latent or not right? | 17:19 |
n4nd0 | really care*, sorry | 17:19 |
wiking | well | 17:20 |
wiking | you can take it like that as well | 17:20 |
n4nd0 | for me it looks more unified in that way, what do you think? | 17:20 |
wiking | the idea i had is just that we make a new feature or whatever for SO | 17:20 |
-!- vikram360 [~vikram360@117.192.189.186] has joined #shogun | 17:20 | |
wiking | and from that i'll inherit latent... to be able to introduce the latent variable | 17:20 |
wiking | because not CLatentFeatures : public CDotFeatures | 17:21 |
wiking | *now | 17:21 |
wiking | s/not/now/g ;) | 17:21 |
n4nd0 | :) | 17:21 |
wiking | so i guess you'd do the same | 17:21 |
wiking | but only have x and y | 17:21 |
wiking | where y can be more than a simple -1,1 | 17:21 |
n4nd0 | why to inherit from CDotFeatures? | 17:22 |
wiking | and you'll need to be able to define the delta(y,y') as well somehow | 17:22 |
wiking | w*PSI(x,y) | 17:22 |
wiking | or in other way <w, PSI(x,y)> | 17:23 |
wiking | so that's a dot product | 17:23 |
n4nd0 | my plan is to have a general class for the labels where that class has a pure virtual method for the cost function (this delta(y, y')) | 17:23 |
wiking | ah yeah | 17:23 |
wiking | that's cool | 17:24 |
n4nd0 | what I have not clear yet is where this class should fit | 17:24 |
wiking | so in case of any application of SO | 17:24 |
n4nd0 | if inheriting from sth else in Shogun or not | 17:24 |
wiking | you'd need to derive from that class and define that pure virtual func | 17:24 |
n4nd0 | exactly | 17:24 |
wiking | well | 17:24 |
wiking | that good but then again | 17:25 |
wiking | you can start the refactoring on the level of CLabels | 17:25 |
wiking | so that CLabels has a delta as well or? :) | 17:25 |
n4nd0 | yes, I wanted to check thoroughly the hierarchy there | 17:25 |
wiking | i mean it's straightforward what is the delta | 17:25 |
n4nd0 | wiking, but it is application dependent again | 17:26 |
n4nd0 | wiking, or at least dependent on the structure used for the labels | 17:27 |
pluskid | that's why we need SequenceLabels, TreeLabels, etc. etc. :p | 17:27 |
n4nd0 | pluskid, those should inherit from this general class labels | 17:28 |
pluskid | yes, did you see the figure and pseudo code? | 17:28 |
n4nd0 | oh, I think I missed that | 17:28 |
pluskid | yuml.me/f14284c4.jpg | 17:29 |
pluskid | and https://gist.github.com/2489859 | 17:29 |
pluskid | very preliminary, though | 17:29 |
pluskid | and there's a problem of using template here: the superclass CLabels is no longer *one* class | 17:29 |
pluskid | but many many classes depending on the template parameter | 17:30 |
pluskid | so maybe we will not use this type of hierarchy, not sure currently | 17:30 |
n4nd0 | pluskid, mmm I don't understand why do you want to have the base class CLabels templated | 17:33 |
pluskid | n4nd0: otherwise, we do not know what to return in the function get_label(int32_t) | 17:33 |
pluskid | it might be int, float, or some tree or sequence | 17:34 |
pluskid | will be complex object in SO I guess | 17:34 |
n4nd0 | pluskid, well but one should know in the moment of using that function | 17:34 |
n4nd0 | should know it* | 17:34 |
pluskid | n4nd0: but how do you define the prototype of that function? | 17:34 |
pluskid | in a common superclass CLabels | 17:35 |
n4nd0 | pluskid, doing something similar to what we have in CFeatures | 17:36 |
pluskid | n4nd0: what's that? | 17:36 |
pluskid | you mean we do not define this function in the base class | 17:36 |
pluskid | but we dynamic_cast to subclasses when calling some specific functions? | 17:37 |
n4nd0 | pluskid, wait a second I am looking for an example within the code so you can see what I mean | 17:37 |
pluskid | ok | 17:37 |
n4nd0 | yes my idea would be to dynamic_cast to subclasses | 17:42 |
blackburn | we don't really need base CLabels | 17:42 |
blackburn | multiclass work with multiclass labels (only) | 17:42 |
pluskid | OK, I understand | 17:42 |
blackburn | binary with binary | 17:42 |
blackburn | however this can be a problem again | 17:42 |
blackburn | cause we have generic machine | 17:43 |
n4nd0 | blackburn, but there might some common functionality to all of them interesting to have in a base class | 17:43 |
pluskid | what is generic machine? | 17:43 |
blackburn | CMachine | 17:43 |
blackburn | it has labels.. | 17:43 |
n4nd0 | blackburn, and I think it offers more flexibility to have a base class | 17:43 |
pluskid | we can have a pure virtual CLabel super class | 17:44 |
pluskid | with almost nothing in it | 17:44 |
n4nd0 | yes | 17:44 |
blackburn | yes but we have to be careful | 17:44 |
n4nd0 | I imagine it like that | 17:44 |
n4nd0 | blackburn, in what sense? | 17:44 |
blackburn | with future casts | 17:44 |
n4nd0 | can you put an example where a problem may arise? | 17:45 |
blackburn | n4nd0: it would be nice to have set_labels(MulticlassLabels) for multiclass machines | 17:45 |
blackburn | but how to do that? | 17:45 |
blackburn | no way probably so it should be checked probably | 17:46 |
pluskid | yes dynamic_cast everywhere | 17:47 |
blackburn | pluskid: n4nd0: wiking: could you please elaborate tihs problem at the mailing list? | 17:48 |
n4nd0 | blackburn, sure, but I have to think more of it first | 17:48 |
blackburn | yes no hurry | 17:48 |
wiking | hehehe ok | 17:49 |
wiking | brb | 17:49 |
pluskid | sure | 17:49 |
blackburn | but we all can forget something so would be nice to have it on the mailing list | 17:49 |
blackburn | thanls | 17:49 |
blackburn | ks* | 17:49 |
n4nd0 | pluskid, blackburn, also, as far as I have understand from SO, SO generalizes multiclass | 17:49 |
pluskid | doesn't shogun has an issue list? | 17:49 |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking] | 17:49 | |
blackburn | n4nd0: yes | 17:49 |
blackburn | pluskid: github issues work well too | 17:49 |
blackburn | feel free to use it | 17:49 |
blackburn | may be even better | 17:49 |
blackburn | yes | 17:49 |
pluskid | blackburn: no body looks at it I guess? :p | 17:49 |
n4nd0 | I think that should be reflected in the diagram somehow | 17:49 |
blackburn | pluskid: will if it is used | 17:50 |
pluskid | n4nd0: in my view, SO is multiclass, with structure in the class space | 17:50 |
blackburn | n4nd0: pluskid: please create new issue - we can even refer code modifications to it | 17:50 |
blackburn | multiclass is SO I'd say | 17:50 |
blackburn | ehmm | 17:50 |
blackburn | subset of SO | 17:50 |
n4nd0 | I agree with blackburn | 17:50 |
blackburn | simple finite | 17:50 |
pluskid | yes, SO without structure haha | 17:51 |
blackburn | no there are some structure | 17:51 |
blackburn | just like structure of N set | 17:51 |
blackburn | there is an order and distance | 17:51 |
pluskid | no, simple discrete set | 17:51 |
pluskid | I don't think there is order in generic multiclass | 17:51 |
blackburn | however d(y,y') = [y=y'] is strange distance | 17:51 |
pluskid | class 1 is not before class 2 | 17:51 |
blackburn | heh right | 17:51 |
blackburn | no reason to keep it in mind | 17:52 |
n4nd0 | there is no really structure in the sense that is important for SO where there are important dependencies between the parts that form a label | 17:52 |
blackburn | I just thought of implementation | 17:52 |
n4nd0 | ok | 17:52 |
blackburn | actually it can be considered as consistent way | 17:52 |
blackburn | to define distance between numbers | 17:52 |
blackburn | like in multiclass | 17:52 |
pluskid | regression can also be thought to be multiclass: with infinite number of classes, and very nice structure (orders, distance, innerproduct, etc.) in the real line :p | 17:53 |
blackburn | brb | 17:53 |
pluskid | midnight here | 17:54 |
pluskid | good night all | 17:54 |
@sonney2k | wtf | 17:54 |
n4nd0 | pluskid, good night | 17:54 |
@sonney2k | you all are going berserk here? | 17:54 |
n4nd0 | :DD | 17:54 |
@sonney2k | I can hardly read the whole log | 17:54 |
@sonney2k | pluskid, indeed no to multiple inheritance | 17:55 |
n4nd0 | I am just preparing by databases exam for tomorrow :P | 17:55 |
@sonney2k | it will complicate things beyond repair | 17:55 |
@sonney2k | and I cannot even think of how one could possibly use swig with that | 17:55 |
pluskid | sonney2k: OK, I'll tell wiking that if he wants to use multiple inheritance, sonney2k will kill him :D | 17:55 |
@sonney2k | (think java, python that all have no MI) | 17:55 |
@sonney2k | I will kill everyone who wants MI | 17:56 |
pluskid | :D | 17:56 |
pluskid | good night | 17:56 |
-!- pluskid [~chatzilla@111.120.57.192] has quit [Quit: ChatZilla 0.9.88.2 [Firefox 12.0/20120420145725]] | 17:56 | |
@sonney2k | pluskid, n8 | 17:56 |
@sonney2k | blackburn, about the label redesign - I am very much in favor of it - but I have no idea how we even can have an interface there | 17:57 |
@sonney2k | I mean basically multiclass, binary and real-valued labels share all functionality | 17:57 |
@sonney2k | only that they accept different values | 17:57 |
n4nd0 | sonney2k, I know that I am on risk of getting killed but ... I think python supports MI | 17:57 |
@sonney2k | like +1/-1 | 17:57 |
@sonney2k | n4nd0, no | 17:57 |
n4nd0 | ok | 17:58 |
n4nd0 | http://docs.python.org/tutorial/classes.html | 17:59 |
n4nd0 | search for multiple inheritance | 17:59 |
n4nd0 | limited form of it they say | 17:59 |
@sonney2k | n4nd0, yeah but look at it - you will more likely shoot yourself in the foot then benefiting from this | 18:00 |
n4nd0 | but it exists :) | 18:00 |
n4nd0 | this is awesome | 18:01 |
n4nd0 | "all cases of multiple inheritance exhibit one or more diamond relationships" | 18:01 |
n4nd0 | :O | 18:01 |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 18:05 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 18:05 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 18:05 | |
-!- n4nd0 [~n4nd0@wifi-cct-cisco-r252-4.cct.urjc.es] has quit [Read error: Connection reset by peer] | 18:12 | |
blackburn | sonney2k: I told you it will be crazy | 18:22 |
-!- vikram360 [~vikram360@117.192.189.186] has quit [Ping timeout: 240 seconds] | 18:49 | |
-!- blackburn [~qdrgsm@188.168.2.179] has quit [Quit: Leaving.] | 18:53 | |
-!- harshit_ [~harshit@182.68.43.52] has joined #shogun | 19:13 | |
-!- harshit_ [~harshit@182.68.43.52] has quit [Client Quit] | 19:14 | |
-!- vikram360 [~vikram360@117.192.173.180] has joined #shogun | 19:19 | |
-!- vikram360 [~vikram360@117.192.173.180] has quit [Ping timeout: 260 seconds] | 19:29 | |
-!- vikram360 [~vikram360@117.192.172.250] has joined #shogun | 19:30 | |
-!- Marty28 [~marty@cable-158-181-78-199.cust.telecolumbus.net] has joined #shogun | 20:00 | |
-!- Marty28 [~marty@cable-158-181-78-199.cust.telecolumbus.net] has quit [Quit: Colloquy for iPad - http://colloquy.mobi] | 20:11 | |
-!- blackburn [5bde8018@gateway/web/freenode/ip.91.222.128.24] has joined #shogun | 20:47 | |
-!- blackburn [5bde8018@gateway/web/freenode/ip.91.222.128.24] has quit [Ping timeout: 245 seconds] | 20:56 | |
@sonney2k | blackburn - all your fault ;-) | 21:31 |
-!- blackburn [5bde8018@gateway/web/freenode/ip.91.222.128.24] has joined #shogun | 21:45 | |
blackburn | sonney2k: in which means? | 21:46 |
@sonney2k | you seem to want to do 1000 things at the same time ;) | 21:58 |
-!- PhilTillet [~Philippe@vir78-1-82-232-38-145.fbx.proxad.net] has quit [Remote host closed the connection] | 21:59 | |
@sonney2k | blackburn, btw I've seen you label class hierarchy | 22:00 |
@sonney2k | which functions should go into the new CLabels? | 22:00 |
blackburn | sonney2k: mine? it was pluskid's I think | 22:05 |
blackburn | sonney2k: this should be discussed actually | 22:06 |
blackburn | about 1000 things - yes but we have enough horsepower^W human resources | 22:07 |
blackburn | so it is a question of management :) | 22:09 |
blackburn | sonney2k: one guy from my school is this year's gsoc student - so a chance I'll get some opencv expert in touch | 22:11 |
@sonney2k | blackburn, because I think what we currently have in CLabels would work for binary/multiclass/reallabels | 22:14 |
@sonney2k | only that we should have some check_consistency / validity of the labels function | 22:15 |
blackburn | sonney2k: I agree but what about structured labels.. | 22:15 |
@sonney2k | they are totally different | 22:15 |
@sonney2k | there is nothing except checking 'is_valid()' that they have in common | 22:16 |
@sonney2k | I mean even get_label() cannot be shared among label classes | 22:16 |
@sonney2k | so it cannot be in the new CLabels() | 22:16 |
@sonney2k | so I am actually not so sure if we really need a CLabels base class | 22:17 |
blackburn | agree | 22:17 |
blackburn | sonney2k: actually it could be shared using templates | 22:18 |
blackburn | but some non-trivial thing | 22:18 |
-!- blackburn [5bde8018@gateway/web/freenode/ip.91.222.128.24] has quit [Ping timeout: 245 seconds] | 22:33 | |
@sonney2k | blackburn - with templates overloading sucks | 22:35 |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking] | 22:39 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 22:46 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 22:46 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 22:46 | |
wiking | dooh | 22:51 |
-!- blackburn [5bde8018@gateway/web/freenode/ip.91.222.128.24] has joined #shogun | 22:51 | |
wiking | https://github.com/shogun-toolbox/shogun/graphs/contributors | 22:51 |
wiking | wonder where am i :) | 22:51 |
wiking | github aaate meee! | 22:53 |
blackburn | that sucks | 22:55 |
blackburn | ;) | 22:55 |
@sonney2k | interesting stats | 22:55 |
wiking | sonney2k: yeah the punchcard is funny as well :) | 22:55 |
blackburn | sonney2k: have you seen my mail to slep? do you think I ask for reasonable things? | 22:56 |
blackburn | sonney2k: around? | 23:01 |
@sonney2k | blackburn, yeah but free for academic use means - not open source | 23:01 |
blackburn | sonney2k: gsomix needs some guidance | 23:01 |
@sonney2k | so you need some statement from them that you can integrate it under the terms of GPL | 23:02 |
blackburn | sonney2k: do you think they would deny it? | 23:02 |
blackburn | ah | 23:02 |
@sonney2k | currently it looks like this | 23:02 |
blackburn | yes so I asked they | 23:02 |
@sonney2k | gsomix, wassup? | 23:02 |
-!- harshit_ [~harshit@182.68.43.52] has joined #shogun | 23:02 | |
blackburn | sonney2k: that would be pretty bad | 23:02 |
gsomix | sonney2k, what we plan to store in objects at end? references to SGVectors or pointers, or copies? | 23:03 |
-!- blackburn_ [5bde8018@gateway/web/freenode/ip.91.222.128.24] has joined #shogun | 23:03 | |
blackburn_ | hmm | 23:03 |
blackburn_ | browser has died | 23:03 |
@sonney2k | gsomix, good question | 23:06 |
blackburn_ | sonney2k: why would they want to keep code available but not real open source? | 23:06 |
@sonney2k | blackburn, why is svmlight/struct not opensource? | 23:06 |
-!- blackburn [5bde8018@gateway/web/freenode/ip.91.222.128.24] has quit [Ping timeout: 245 seconds] | 23:07 | |
blackburn_ | sonney2k: isn't it? | 23:07 |
@sonney2k | of course not | 23:07 |
blackburn_ | that's crazy | 23:07 |
@sonney2k | blackburn_, that is why svmlight is not part of official shogun | 23:08 |
blackburn_ | sonney2k: yes I know it has some special status | 23:08 |
blackburn_ | can SLEP has similar one then? | 23:08 |
@sonney2k | and I was begging TJ for way too many times - he simply doesn't like OSS | 23:08 |
@sonney2k | blackburn_, ask them: tell them that shogun is under GPLv3 and that you need a statement saying SLEP is released under a GPL3 compatible license or under GPL3 directly | 23:09 |
blackburn_ | sonney2k: can reimplementation of anything has different license? | 23:10 |
@sonney2k | gsomix, let me think loudly: If we were using pointers we would keep ptrs everywhere. | 23:10 |
blackburn_ | sonney2k: yeah I asked already but not about statement - lets wait for answer first | 23:10 |
@sonney2k | blackburn_, re-implementing without using / looking at their code - of course | 23:10 |
blackburn_ | sonney2k: that license stuff is pretty vomitting isn't it? | 23:11 |
@sonney2k | blackburn_, sure it is - in particular because even OSS licenses might conflict with each other - so we cannot use MPL / CPL licensed code in shogun | 23:12 |
@sonney2k | gsomix, so in principle we want exactly that - pass around object ptrs only | 23:12 |
blackburn_ | lets hope they will license it using gpl | 23:13 |
blackburn_ | I think to allow to be integrated to shogun is the only way to survive :D | 23:17 |
@sonney2k | gsomix, now when we have a function SGVector return_new_vector() we have to return some copy | 23:17 |
@sonney2k | gsomix, but otherwise we return the reference SGVector& return_vector_from_class() | 23:18 |
-!- blackburn_ [5bde8018@gateway/web/freenode/ip.91.222.128.24] has quit [Ping timeout: 245 seconds] | 23:21 | |
gsomix | sonney2k, aha. | 23:24 |
gsomix | sonney2k, sonney2k, btw, what about `const SGVector& return_vector_from_class() const;'? Should we have two version (one for const objects)? | 23:24 |
gsomix | sonney2k, ok, I'll think about it yet. I will end review of code for `SGVector to refs' tomorrow. | 23:37 |
gsomix | I'm going to bed now =___= | 23:38 |
gsomix | nite | 23:38 |
-!- blackburn_ [5bde8018@gateway/web/freenode/ip.91.222.128.24] has joined #shogun | 23:49 | |
blackburn_ | errr connection sucks | 23:49 |
blackburn_ | when I see mails like that last at the gsoc students mailing list | 23:50 |
blackburn_ | I ask myself how can people be that stupid | 23:50 |
@sonney2k | blackburn_, I start to wonder if we shouldn't use SGVector* ... | 23:54 |
@sonney2k | I feel uneasy by returning copies of SGVector and not being able to store references... | 23:55 |
blackburn_ | sonney2k: I am afraid we can stuck there | 23:55 |
@sonney2k | what do you mean? | 23:55 |
blackburn_ | sonney2k: I mean we can lose much time doing this (being unsure it is useful) | 23:56 |
blackburn_ | I am talking about any transitions | 23:56 |
@sonney2k | blackburn_, sure | 23:56 |
@sonney2k | that is why I am asking | 23:57 |
@sonney2k | blackburn_, btw why do you use the webchat? | 23:57 |
blackburn_ | sonney2k: using public wifi - irc is blocked | 23:57 |
blackburn_ | sonney2k: we would need to really get sure we need it | 23:58 |
blackburn_ | do we have any problem with it so far? | 23:58 |
@sonney2k | yes | 23:58 |
blackburn_ | which one? | 23:58 |
@sonney2k | everything is copied and we have no proper way of cleaning up memory | 23:58 |
--- Log closed Thu Apr 26 00:00:01 2012 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!