--- Log opened Tue May 13 00:00:46 2014 | ||
-!- lisitsyn [~qdrgsm@80.252.20.67] has left #shogun [] | 00:15 | |
-!- [1]Saurabh7 [~Saurabh7@115.248.45.78] has joined #shogun | 00:21 | |
-!- Saurabh7 [~Saurabh7@115.248.45.78] has quit [Ping timeout: 265 seconds] | 00:24 | |
-!- [1]Saurabh7 is now known as Saurabh7 | 00:24 | |
-!- [1]Saurabh7 [~Saurabh7@115.248.45.78] has joined #shogun | 00:28 | |
-!- Saurabh7 [~Saurabh7@115.248.45.78] has quit [Ping timeout: 276 seconds] | 00:31 | |
-!- [1]Saurabh7 is now known as Saurabh7 | 00:31 | |
-!- DarkRoom [~darkroom@14.139.82.6] has joined #shogun | 01:10 | |
-!- PirosB3 [~pirosb3@cpc4-newt31-2-0-cust631.newt.cable.virginm.net] has quit [Quit: PirosB3] | 01:37 | |
-!- PirosB3 [~pirosb3@cpc4-newt31-2-0-cust631.newt.cable.virginm.net] has joined #shogun | 01:40 | |
-!- pickle27 [~pickle27@192-0-136-118.cpe.teksavvy.com] has joined #shogun | 01:49 | |
-!- pickle27 [~pickle27@192-0-136-118.cpe.teksavvy.com] has quit [Client Quit] | 01:49 | |
-!- pickle27 [~pickle27@192-0-136-118.cpe.teksavvy.com] has joined #shogun | 01:49 | |
-!- PirosB3 [~pirosb3@cpc4-newt31-2-0-cust631.newt.cable.virginm.net] has quit [Quit: PirosB3] | 02:08 | |
-!- Saurabh7 [~Saurabh7@115.248.45.78] has quit [Ping timeout: 240 seconds] | 05:31 | |
-!- pickle27 [~pickle27@192-0-136-118.cpe.teksavvy.com] has quit [Remote host closed the connection] | 05:41 | |
-!- kislay [~Abhijeet@103.240.205.246] has quit [Ping timeout: 250 seconds] | 05:52 | |
-!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has quit [Remote host closed the connection] | 06:12 | |
-!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has joined #shogun | 06:14 | |
-!- iglesiasg [~iglesiasg@524AE0A7.cm-4-3d.dynamic.ziggo.nl] has joined #shogun | 06:21 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 06:21 | |
@iglesiasg | Good morning people | 06:22 |
---|---|---|
-!- parijat [671b082a@gateway/web/freenode/ip.103.27.8.42] has joined #shogun | 06:58 | |
@iglesiasg | parijat, Hey parijat, I'm having a look at your PR atm | 07:00 |
parijat | iglesiasg: hi.. i see that you are reviewing my PR.. if you have any questions, I am available here in irc.. | 07:00 |
@iglesiasg | parijat, great, thanks | 07:00 |
parijat | iglesiasg: yup I saw! | 07:00 |
@iglesiasg | parijat, for the moment all of them are more or less minor comments | 07:02 |
@iglesiasg | parijat, but please have a look at the comment on line 152 of C45ClassifierTree | 07:02 |
parijat | iglesiasg: yup, I saw a few of them.. I will address them | 07:02 |
@iglesiasg | parijat, maybe I misunderstood something of what's going on there | 07:02 |
parijat | iglesiasg: alright,, let me see. | 07:02 |
@iglesiasg | parijat, not exactly on that line, pretty close | 07:03 |
parijat | iglesiasg: I think you are right there! if the last set of same labels has higher weight, its gooing to ignore that. | 07:05 |
@iglesiasg | parijat, all right | 07:05 |
parijat | iglesiasg: I will fix this | 07:05 |
@iglesiasg | parijat, I think you can correct if by pulling out the else if and making it an independent if within the for loop | 07:05 |
parijat | iglesiasg: yes exactly! | 07:06 |
@iglesiasg | parijat, I compiled and tested your code locally btw | 07:09 |
@iglesiasg | parijat, 0 leaks in valgrind :) | 07:09 |
parijat | iglesiasg: me too, I checked valgrind before sending in! ;) | 07:10 |
parijat | iglesiasg: btw, did you get any word about acceptance of your paper? | 07:13 |
@iglesiasg | parijat, yes, yesterday night | 07:14 |
@iglesiasg | parijat, they accepted it \o/ | 07:14 |
@iglesiasg | parijat, but I have to address some comments from reviewers | 07:14 |
parijat | iglesiasg: yo! calls for celebrations. | 07:15 |
@iglesiasg | parijat, hehe yeah! | 07:15 |
parijat | iglesiasg: very nice! so you would be travelling somewhere to present the paper? | 07:15 |
@iglesiasg | parijat, haha yeah | 07:16 |
@iglesiasg | to Spain.... | 07:16 |
@iglesiasg | hahaha | 07:16 |
@iglesiasg | funny thing | 07:16 |
@iglesiasg | at least it is not to my city, although another city I've already been in. Salamanca | 07:16 |
@iglesiasg | parijat, this if the conference: http://fusion2014.org/ | 07:17 |
parijat | iglesiasg: haha! good good.. | 07:17 |
@iglesiasg | I'd prefer to visit a new place | 07:17 |
parijat | iglesiasg: oh! let me see.. | 07:17 |
parijat | iglesiasg: ya, Im sure you will get many opportunities to visit new places, this is just the beginning.. | 07:18 |
parijat | iglesiasg: you are right about data getting modified... I should probably restore the data matrix | 07:21 |
-!- kislay [~Abhijeet@103.240.206.190] has joined #shogun | 07:27 | |
@iglesiasg | parijat, ok, I should start going soon | 07:28 |
@iglesiasg | parijat, I have left to look into entropy, informational_gain... and apply_multiclass_from_... | 07:29 |
@iglesiasg | parijat, but I guess these are pretty much the same as in ID3, right? | 07:29 |
parijat | iglesiasg: ya almost! very few changes | 07:29 |
@iglesiasg | parijat, I should be able to read them during the day anyway | 07:30 |
@iglesiasg | parijat, nice job! :) | 07:30 |
parijat | iglesiasg: thanks! there are a few things that are yet to be done like missing values.. | 07:31 |
parijat | iglesiasg: should I do these in this PR itself? or is it better to wait for this getting merged? | 07:31 |
@iglesiasg | parijat, better wait, this PR is already large | 07:32 |
@iglesiasg | parijat, remind me what these missing values are, please | 07:32 |
parijat | iglesiasg: i meant feature for handling missing values in C4.5 | 07:32 |
parijat | iglesiasg: i have to add this. plus tree pruning | 07:33 |
parijat | iglesiasg: but this PR is complete in itself without these. So it wont be a problem if we merge this 1st and then go ahead adding the remaining things | 07:34 |
@iglesiasg | parijat, I meant what the idea was behind handling missing values | 07:37 |
@iglesiasg | parijat, I have bad memory :) | 07:37 |
parijat | iglesiasg: ok! the idea is that.. | 07:38 |
parijat | iglesiasg: if we choose a data vector with unknown chosen attribute, we split the entire weight of the vector into fractions correponding to the possibility of the vector being moved to each of the child nodes | 07:39 |
parijat | iglesiasg: so basically we move the vector into all the children nodes of the current node, but the weight gets divided accordingly | 07:40 |
parijat | iglesiasg: did I make sense? | 07:42 |
@iglesiasg | parijat, yep | 07:43 |
parijat | iglesiasg: sometimes what I say can be difficult to understand!! :) | 07:44 |
@iglesiasg | parijat, if the attribute is unknown, how are the weights computed in each of the children? | 07:45 |
@iglesiasg | parijat, the same for all of them? | 07:45 |
@iglesiasg | I mean these possibilities you mentioned above | 07:45 |
parijat | iglesiasg: no, based on the ratio in which vectors with know attributes move into various children | 07:45 |
@iglesiasg | parijat, ok | 07:46 |
@iglesiasg | thanks | 07:46 |
parijat | iglesiasg: not a problem! :) | 07:47 |
@iglesiasg | parijat, this is in some way semi-supervised learning, isn't it? | 07:47 |
parijat | iglesiasg: ya, in a way.. but not entirely though I think.. its just extended to handle missing values | 07:48 |
parijat | iglesiasg: I too have to go now.. let me address your comments and update the PR | 07:53 |
-!- parijat [671b082a@gateway/web/freenode/ip.103.27.8.42] has quit [Quit: Page closed] | 07:57 | |
@iglesiasg | catch you later | 08:00 |
-!- iglesiasg [~iglesiasg@524AE0A7.cm-4-3d.dynamic.ziggo.nl] has quit [Quit: Leaving] | 08:00 | |
-!- PirosB3 [~pirosb3@cpc4-newt31-2-0-cust631.newt.cable.virginm.net] has joined #shogun | 09:46 | |
-!- kislay [~Abhijeet@103.240.206.190] has quit [Ping timeout: 240 seconds] | 09:49 | |
-!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun | 10:05 | |
-!- kislay [~Abhijeet@103.240.205.72] has joined #shogun | 10:10 | |
-!- kislay [~Abhijeet@103.240.205.72] has quit [Ping timeout: 258 seconds] | 10:15 | |
-!- thoralf [~thoralf@91-66-177-150-dynip.superkabel.de] has joined #shogun | 10:18 | |
thoralf | Hellau. | 10:20 |
lisitsyn | thoralf: today I am talking about 1+2+3+... = -1/12 | 10:34 |
lisitsyn | what do you think? | 10:34 |
thoralf | lisitsyn: True if "... = -73/12" | 10:38 |
lisitsyn | thoralf: nah not really ;) | 10:38 |
thoralf | of course! | 10:38 |
thoralf | lisitsyn: Tell me - what is it about? | 11:00 |
lisitsyn | thoralf: http://en.wikipedia.org/wiki/1_%2B_2_%2B_3_%2B_4_%2B_%E2%8B%AF | 11:01 |
thoralf | Crazy | 11:03 |
-!- soumyaC [uid15286@gateway/web/irccloud.com/x-glhceqzfxnrglcic] has joined #shogun | 11:27 | |
-!- kislay [~Abhijeet@103.240.205.135] has joined #shogun | 11:30 | |
-!- kislay [~Abhijeet@103.240.205.135] has quit [Ping timeout: 252 seconds] | 11:36 | |
-!- HeikoS [~heiko@0541877d.skybroadband.com] has joined #shogun | 12:25 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 12:25 | |
-!- kislay [~Abhijeet@103.240.206.95] has joined #shogun | 13:00 | |
-!- PirosB3 [~pirosb3@cpc4-newt31-2-0-cust631.newt.cable.virginm.net] has quit [Quit: PirosB3] | 13:32 | |
-!- parijat [671b082a@gateway/web/freenode/ip.103.27.8.42] has joined #shogun | 14:06 | |
-!- HeikoS [~heiko@0541877d.skybroadband.com] has quit [Ping timeout: 240 seconds] | 14:18 | |
-!- PirosB3 [~pirosb3@cpc4-newt31-2-0-cust631.newt.cable.virginm.net] has joined #shogun | 14:20 | |
-!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.] | 14:42 | |
-!- HeikoS [~heiko@pat-191-250.internal.eduroam.ucl.ac.uk] has joined #shogun | 15:41 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 15:41 | |
-!- pickle27 [~pickle27@192-0-136-118.cpe.teksavvy.com] has joined #shogun | 15:45 | |
kislay | hey pickle27 :) | 16:12 |
-!- pickle27 [~pickle27@192-0-136-118.cpe.teksavvy.com] has quit [Remote host closed the connection] | 16:13 | |
-!- pickle27f [~pickle27@216.191.231.218] has joined #shogun | 16:23 | |
kislay | hey pickle27 :) | 16:36 |
kislay | pickle27f, looks like some network problems you are suffering with! | 16:36 |
pickle27f | kislay, agh yeah this is my work computer | 16:37 |
pickle27f | macs are the worst at losing wifi and being weird about it | 16:37 |
kislay | pickle27f, haha :D | 16:37 |
kislay | pickle27f, I have been reading quite a lot about factory pattern! | 16:37 |
kislay | pickle27f, but there seems too unorganized content present! | 16:38 |
pickle27f | thats great - yeah don't over do it, you can probably start implementing the shogun opencv converter factories now | 16:40 |
kislay | oh. yeah . will love to! | 16:41 |
pickle27f | also can you take a look at the PR - we need to put OpenCV under a flag | 16:41 |
kislay | yes. | 16:42 |
kislay | I was about to look into it | 16:42 |
-!- PirosB3 [~pirosb3@cpc4-newt31-2-0-cust631.newt.cable.virginm.net] has quit [Quit: PirosB3] | 16:42 | |
-!- PirosB3 [~pirosb3@cpc4-newt31-2-0-cust631.newt.cable.virginm.net] has joined #shogun | 16:46 | |
-!- parijat [671b082a@gateway/web/freenode/ip.103.27.8.42] has quit [Quit: Page closed] | 17:14 | |
-!- HeikoS [~heiko@pat-191-250.internal.eduroam.ucl.ac.uk] has quit [Ping timeout: 240 seconds] | 17:22 | |
-!- HeikoS [~heiko@pat-191-250.internal.eduroam.ucl.ac.uk] has joined #shogun | 17:23 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 17:23 | |
kislay | pickle27f, I couldn't understand the reason we need a flag. you mentioned there that if someone wants OpenCV integration and they don't have OPenCV installed, we need not give them warning. But why? | 17:36 |
pickle27f | we want OpenCV integration to be something the user specifies when configuring shogun | 17:37 |
pickle27f | the same way the user specifies if they want python | 17:37 |
pickle27f | we don't simply build python support if we can | 17:37 |
pickle27f | we only do it if asked | 17:37 |
pickle27f | we want the same for OpenCV | 17:37 |
pickle27f | so it needs to be behind a flag | 17:37 |
pickle27f | does that make sense? | 17:37 |
kislay | pickle27f, yup. | 17:38 |
pickle27f | that way the default build does not change at all and if they want OpenCV we'll try and add it | 17:38 |
pickle27f | if they want opencv but don't have it installed we'll say something about it | 17:38 |
kislay | oh! yeah | 17:38 |
kislay | okay. got it. | 17:39 |
-!- HeikoS [~heiko@pat-191-250.internal.eduroam.ucl.ac.uk] has quit [Quit: Leaving.] | 18:10 | |
pickle27f | guys why are my emails to shogun-team always filtered? can I be whitelisted on there | 18:18 |
pickle27f | sonney2k_, ^ | 18:18 |
-!- pickle27f [~pickle27@216.191.231.218] has quit [Quit: This computer has gone to sleep] | 18:37 | |
-!- thoralf [~thoralf@91-66-177-150-dynip.superkabel.de] has quit [Quit: Konversation terminated!] | 18:39 | |
-!- pickle27f [~pickle27@216.191.231.218] has joined #shogun | 18:56 | |
-!- HeikoS [~heiko@0541877d.skybroadband.com] has joined #shogun | 19:15 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 19:15 | |
-!- HeikoS [~heiko@0541877d.skybroadband.com] has quit [Ping timeout: 255 seconds] | 19:39 | |
-!- pickle27f [~pickle27@216.191.231.218] has quit [Quit: Leaving] | 19:44 | |
-!- HeikoS [~heiko@0541877d.skybroadband.com] has joined #shogun | 19:59 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 19:59 | |
-!- HeikoS [~heiko@0541877d.skybroadband.com] has quit [Quit: Leaving.] | 20:08 | |
-!- soumyaC [uid15286@gateway/web/irccloud.com/x-glhceqzfxnrglcic] has quit [Quit: Connection closed for inactivity] | 20:26 | |
-!- utsavjain [~utsav@113.193.139.79] has joined #shogun | 20:36 | |
-!- PirosB3 [~pirosb3@cpc4-newt31-2-0-cust631.newt.cable.virginm.net] has quit [Quit: PirosB3] | 20:45 | |
-!- PirosB3 [~pirosb3@cpc4-newt31-2-0-cust631.newt.cable.virginm.net] has joined #shogun | 21:01 | |
-!- pickle27f [~pickle27@216.191.231.218] has joined #shogun | 21:16 | |
-!- utsavjain [~utsav@113.193.139.79] has left #shogun [] | 21:33 | |
-!- pickle27f [~pickle27@216.191.231.218] has quit [Quit: This computer has gone to sleep] | 21:39 | |
-!- pickle27f [~pickle27@216.191.231.218] has joined #shogun | 21:42 | |
-!- HeikoS [~heiko@0541877d.skybroadband.com] has joined #shogun | 22:57 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 22:58 | |
-!- iglesiasg [~iglesiasg@524AE0A7.cm-4-3d.dynamic.ziggo.nl] has joined #shogun | 23:11 | |
-!- mode/#shogun [+o iglesiasg] by ChanServ | 23:12 | |
@iglesiasg | sup guys | 23:12 |
-!- pickle27f [~pickle27@216.191.231.218] has quit [Quit: This computer has gone to sleep] | 23:26 | |
-!- kislay [~Abhijeet@103.240.206.95] has quit [Ping timeout: 252 seconds] | 23:35 | |
-!- lisitsyn [~lisitsyn@80.252.20.67] has joined #shogun | 23:52 | |
lisitsyn | hey iglesiasg HeikoS | 23:53 |
lisitsyn | cool news :) | 23:53 |
@iglesiasg | lisitsyn, yandex? | 23:54 |
lisitsyn | no i mean europython thing fonally | 23:55 |
lisitsyn | finally | 23:55 |
@iglesiasg | aaah ok | 23:55 |
lisitsyn | argh i am like hot now :d | 23:55 |
lisitsyn | around 38c you know | 23:56 |
@iglesiasg | ups | 23:57 |
@iglesiasg | lisitsyn, you are probably the warmest part of Russia now | 23:58 |
lisitsyn | haha | 23:58 |
lisitsyn | no its hot outside now | 23:58 |
lisitsyn | so maybe there is a place | 23:58 |
--- Log closed Wed May 14 00:00:26 2014 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!