--- Log opened Tue Jun 17 00:00:36 2014 | ||
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has joined #shogun | 00:07 | |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 00:38 | |
-!- lambday [67157e4d@gateway/web/freenode/ip.103.21.126.77] has joined #shogun | 00:42 | |
-!- zxtx_ [~zv@f052018195.adsl.alicedsl.de] has joined #shogun | 00:45 | |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 00:45 | |
shogun-notifier- | shogun-data: lambday :master * c216ce9 / testsuite/ (8 files): https://github.com/shogun-toolbox/shogun-data/commit/c216ce9b2ed0e668acd7f0b18e5c6fcc6fcd2612 | 00:45 |
---|---|---|
shogun-notifier- | shogun-data: updated data for PR #2309 | 00:45 |
shogun-notifier- | shogun-data: Soumyajit De :master * d3a75a6 / testsuite/ (8 files): https://github.com/shogun-toolbox/shogun-data/commit/d3a75a6ff44b83e173384acc21f5e3cc8e629c52 | 00:45 |
shogun-notifier- | shogun-data: Merge pull request #55 from lambday/master | 00:45 |
shogun-notifier- | shogun-data: | 00:45 |
shogun-notifier- | shogun-data: updated data for PR #2309 | 00:45 |
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has quit [Quit: PirosB3] | 00:45 | |
-!- abinash_panda [~abinash@223.176.23.125] has quit [Ping timeout: 244 seconds] | 00:51 | |
-!- abinash_panda [~abinash@182.66.31.160] has joined #shogun | 00:58 | |
-!- pickle27 [~pickle27@192-0-136-118.cpe.teksavvy.com] has joined #shogun | 01:22 | |
-!- pickle27 [~pickle27@192-0-136-118.cpe.teksavvy.com] has quit [Quit: Leaving] | 01:35 | |
shogun-notifier- | shogun: lambday :feature/mmd * 32f26d6 / / (11 files): https://github.com/shogun-toolbox/shogun/commit/32f26d6fe6a6520f85d7c8baaa1c30e2ad872e86 | 02:19 |
shogun-notifier- | shogun: Fixed covariance computation in CStreamingMMD and some other changes | 02:19 |
shogun-notifier- | shogun: | 02:19 |
shogun-notifier- | shogun: Changes made | 02:19 |
shogun-notifier- | shogun: | 02:19 |
shogun-notifier- | shogun: - While computing covariance matrix in streaming mmd for multiple | 02:19 |
shogun-notifier- | shogun: Soumyajit De :feature/mmd * e38ca25 / / (11 files): https://github.com/shogun-toolbox/shogun/commit/e38ca25ecbbb36839432b2c2ccd3e3dc834c0e0e | 02:19 |
shogun-notifier- | shogun: Merge pull request #2309 from lambday/feature/btest | 02:19 |
shogun-notifier- | shogun: | 02:19 |
shogun-notifier- | shogun: Fixed covariance computation in CStreamingMMD and some other changes | 02:19 |
-!- lambday [67157e4d@gateway/web/freenode/ip.103.21.126.77] has quit [Quit: Leaving.] | 02:36 | |
-!- travis-ci [~travis-ci@ec2-54-83-107-201.compute-1.amazonaws.com] has joined #shogun | 02:49 | |
travis-ci | [travis-ci] it's Soumyajit De's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/27732296 | 02:49 |
-!- travis-ci [~travis-ci@ec2-54-83-107-201.compute-1.amazonaws.com] has left #shogun [] | 02:49 | |
shogun-buildbot | build #30 of nightly_fedora is complete: Failure [failed git] Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_fedora/builds/30 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com>, khalednasr <k.nasr92@gmail.com> | 03:00 |
-!- soumyaC [uid15286@gateway/web/irccloud.com/x-rsvcwvepjgaihcpc] has quit [Quit: Connection closed for inactivity] | 03:16 | |
-!- kislay [~Abhijeet@103.240.205.204] has quit [Ping timeout: 272 seconds] | 03:18 | |
-!- khalednasr [~k.nasr92@41.69.237.204] has quit [Quit: Leaving] | 03:37 | |
-!- wiking_ [~wiking@info2k1.hu] has joined #shogun | 04:11 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Read error: Connection reset by peer] | 04:11 | |
-!- wiking_ [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 04:13 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 04:16 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 04:19 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 04:21 | |
-!- wiking [~wiking@info2k1.hu] has quit [Ping timeout: 264 seconds] | 04:25 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 04:26 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 04:29 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 04:31 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 04:33 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 04:36 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 04:44 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 04:48 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 04:50 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 04:53 | |
-!- wiking_ [~wiking@info2k1.hu] has joined #shogun | 04:58 | |
-!- wiking_ [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 04:58 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 04:58 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 05:03 | |
-!- wiking_ [~wiking@info2k1.hu] has joined #shogun | 05:08 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 05:08 | |
-!- wiking_ [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 05:08 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 05:13 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 05:16 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 05:18 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 05:18 | |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 05:19 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 05:23 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 05:27 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 05:28 | |
-!- wiking [~wiking@info2k1.hu] has quit [Ping timeout: 264 seconds] | 05:33 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 05:33 | |
-!- abinash_panda [~abinash@182.66.31.160] has quit [Ping timeout: 252 seconds] | 05:41 | |
-!- abinash_panda [~abinash@117.200.239.15] has joined #shogun | 05:48 | |
-!- wiking_ [~wiking@info2k1.hu] has joined #shogun | 06:00 | |
-!- wiking [~wiking@info2k1.hu] has quit [Ping timeout: 264 seconds] | 06:03 | |
-!- wiking_ [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 06:03 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 06:05 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 06:10 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 06:10 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 06:10 | |
abinash_panda | /quit | 06:13 |
-!- abinash_panda [~abinash@117.200.239.15] has quit [Quit: leaving] | 06:14 | |
-!- abinash_panda [~abinash@117.200.239.15] has joined #shogun | 06:14 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 06:15 | |
-!- wiking_ [~wiking@info2k1.hu] has joined #shogun | 06:20 | |
-!- wiking [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 06:20 | |
-!- wiking_ [~wiking@info2k1.hu] has quit [Read error: Connection reset by peer] | 06:23 | |
-!- wiking [~wiking@info2k1.hu] has joined #shogun | 06:25 | |
-!- kislay [~Abhijeet@103.240.206.209] has joined #shogun | 08:50 | |
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has joined #shogun | 09:37 | |
shogun-buildbot | build #125 of osx2 - python is complete: Failure [failed test python modular] Build details are at http://buildbot.shogun-toolbox.org/builders/osx2%20-%20python/builds/125 blamelist: Abhijeet <abhijeetkislay@gmail.com> | 10:03 |
-!- Netsplit *.net <-> *.split quits: abinash_panda, wiking, naywhayare, zxtx, sonne|work | 10:16 | |
-!- Netsplit over, joins: wiking, abinash_panda, zxtx, naywhayare, sonne|work | 10:21 | |
-!- wiking_ [~wiking@info2k1.hu] has joined #shogun | 10:24 | |
-!- abinash_panda [~abinash@117.200.239.15] has quit [Write error: Broken pipe] | 10:28 | |
-!- wiking [~wiking@info2k1.hu] has quit [Write error: Broken pipe] | 10:28 | |
shogun-buildbot | build #126 of osx2 - python is complete: Failure [failed test python modular] Build details are at http://buildbot.shogun-toolbox.org/builders/osx2%20-%20python/builds/126 blamelist: lambday <heavensdevil6909@gmail.com>, abinashpanda <abinash.panda.ece10@itbhu.ac.in>, Bj?rn Esser <bjoern.esser@gmail.com>, Parijat Mazumdar <mazumdarparijat@gmail.com>, Kevin Hughes | 10:28 |
shogun-buildbot | <kevinhughes27@gmail.com>, jiaolong <jiaolongxu@gmail.com>, Wu Lin <yorker.lin@gmail.com>, tklein23 <tklein23@users.noreply.github.com>, Soumyajit De <heavensdevil6909@gmail.com>, Thoralf Klein <thoralf@fischlustig.de>, Heiko Strathmann <heiko.strathmann@gmail.com>, Sergey Lisitsyn <lisitsyn.s.o@gmail.com>, Fernando Iglesias <fernando.iglesiasg@gmail.com>, Abhijeet | 10:28 |
shogun-buildbot | <abhijeetkislay@gmail.com>, khalednasr <k.nasr92@gmail.com> | 10:28 |
-!- abinash_panda [~abinash@117.200.239.15] has joined #shogun | 10:29 | |
-!- Netsplit *.net <-> *.split quits: abinash_panda, naywhayare, zxtx | 10:33 | |
-!- Netsplit *.net <-> *.split quits: sonne|work | 10:33 | |
-!- Netsplit over, joins: abinash_panda, zxtx, sonne|work, naywhayare | 10:44 | |
-!- soumyaC [uid15286@gateway/web/irccloud.com/x-rbrtnlwwswhiwekg] has joined #shogun | 10:46 | |
-!- visof [~visof@41.236.191.58] has joined #shogun | 11:24 | |
visof | Hello | 11:24 |
visof | i'm trying to make word2vec | 11:25 |
visof | and i got word2vec.c:1: error: CPU you selected does not support x86-64 instruction set | 11:25 |
visof | any suggestion | 11:25 |
visof | ? | 11:25 |
-!- zxtx_ [~zv@f052018195.adsl.alicedsl.de] has quit [Ping timeout: 272 seconds] | 11:49 | |
wiking_ | word2vec.c ? | 11:57 |
visof | wiking_: yeah | 12:07 |
wiking_ | what's that? | 12:11 |
-!- wiking_ is now known as wiking | 12:11 | |
-!- wiking [~wiking@info2k1.hu] has quit [Changing host] | 12:11 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 12:11 | |
-!- mode/#shogun [+o wiking] by ChanServ | 12:11 | |
-!- visof_ [~visof@unaffiliated/visof] has joined #shogun | 12:15 | |
-!- visof [~visof@41.236.191.58] has quit [Ping timeout: 255 seconds] | 12:15 | |
-!- zxtx_ [~zv@f052018195.adsl.alicedsl.de] has joined #shogun | 13:30 | |
-!- abinash_panda [~abinash@117.200.239.15] has quit [Ping timeout: 240 seconds] | 13:31 | |
-!- PirosB3 [~pirosb3@host80-223-dynamic.246-95-r.retail.telecomitalia.it] has quit [Quit: PirosB3] | 13:37 | |
-!- kislay [~Abhijeet@103.240.206.209] has quit [Ping timeout: 252 seconds] | 14:31 | |
-!- kislay [~Abhijeet@103.240.205.138] has joined #shogun | 14:45 | |
-!- khalednasr [~k.nasr92@41.69.211.87] has joined #shogun | 15:09 | |
-!- lambday [67157f4c@gateway/web/freenode/ip.103.21.127.76] has joined #shogun | 15:18 | |
visof_ | Hell | 15:19 |
visof_ | Hello | 15:19 |
visof_ | is there anybody around? | 15:19 |
lambday | visof_: yes.. hi! | 15:19 |
visof_ | lambday: have you played with freebase before? | 15:21 |
visof_ | or semantic web | 15:21 |
lambday | visof_: um... nope! sorry :) | 15:22 |
visof_ | lambday: have used word2vec? | 15:22 |
visof_ | lambday: if i can two words like "goelocation_latitude" and "latitude" , how can get they are somehow the same thing | 15:23 |
visof_ | maybe also "location_geolocation_latitude" and "latitude" also and comparison like this | 15:24 |
lambday | visof_: I'm afraid I'm not the right person to address this... this is hardcore NLP stuff :/ I don't know much about that | 15:25 |
visof_ | lambday: you don't use word2vec? | 15:25 |
lambday | visof_: no man! sorry :( | 15:26 |
visof_ | haha may i ask why are you here? | 15:26 |
lambday | visof_: I work in a different domain | 15:27 |
visof_ | okay thanks | 15:27 |
lisitsyn | haha I am confused | 15:29 |
lisitsyn | visof_: why do you think we know about word2vec? ;) | 15:29 |
lambday | lisitsyn: :D | 15:29 |
-!- HeikoS [~heiko@nat-240-155.internal.eduroam.ucl.ac.uk] has joined #shogun | 15:43 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 15:43 | |
lambday | lisitsyn: nystisil: both avatars present, eh? | 15:43 |
lisitsyn | lambday: yeah :D | 15:44 |
lambday | man Masters thesis is such a PITA :( | 15:47 |
-!- visof_ is now known as visof | 15:48 | |
lambday | wish I can finish BAHSIC this week.. will get to use it for my thesis :/ | 15:48 |
lambday | HeikoS: hey | 15:48 |
@HeikoS | lambday: hey! | 15:49 |
@HeikoS | lambday: BAHSIC yeah! :) | 15:49 |
@HeikoS | lambday: if you want to, start with that | 15:49 |
lambday | HeikoS: I found another bug in linear time mmd code... fixed it | 15:49 |
-!- visof is now known as weqewewqew | 15:49 | |
@HeikoS | lambday: just had a chat with Dino, who is not about to review the stuff you did so far a bit | 15:49 |
lambday | HeikoS: not about to? | 15:50 |
lisitsyn | lambday: do masters in russia :D | 15:50 |
lisitsyn | easy hah | 15:50 |
@HeikoS | now | 15:50 |
lambday | lisitsyn: yeah.. I should have tried in europe for masters | 15:50 |
@HeikoS | lambday: ^ | 15:50 |
lambday | HeikoS: oh... :D btw did you happen to check the design for BAHSIC? | 15:51 |
lambday | HeikoS: for three variable interaction, I think we need a separate class.. | 15:51 |
lambday | HeikoS: so for three-variable interaction... we'll check two of the features and the labels, right? x_1 and x_2 from p and y from q? | 15:52 |
@HeikoS | lambday: I guess so, I am about to discuss it with dino also | 15:52 |
@HeikoS | will do later today | 15:52 |
lambday | HeikoS: alright.. I also sent b-test patch.. | 15:52 |
@HeikoS | lambday: cool | 15:52 |
@HeikoS | lambday: now back to life as exams/deadline is over | 15:52 |
lambday | HeikoS: travis complains about a libshogun example taking too much time... | 15:52 |
lambday | HeikoS: haha.. yeah and now my deadlines begin.. both here in gsoc and in my insti.. :/ | 15:53 |
@HeikoS | lambday: ah well, you already did lots of things! | 15:54 |
@HeikoS | lambday: but lets push the feature selection stuff. if its useful for your thesis that would be amazinh | 15:54 |
@HeikoS | lambday: what are the immediate next steps? | 15:54 |
lambday | HeikoS: yeah.. so.. I think if we can finalize the design for BAHSIC I can start with that... | 15:54 |
@HeikoS | lambday: let me talk to dino now then | 15:55 |
@HeikoS | could you send an email to both of us with the latest design? | 15:55 |
@HeikoS | (in case things changed?) | 15:55 |
lambday | HeikoS: well, I didn't change anything in the previous design... | 15:55 |
@HeikoS | ok well discuss this one then | 15:55 |
@HeikoS | give me 30 mins | 15:55 |
lambday | HeikoS: thanks :) no hurries though | 15:56 |
@HeikoS | lambday: things for your list also: b-test API exmple, and notebook section | 15:56 |
lambday | HeikoS: yeah... so in the notebook do you think we should use new implementation... or a comparison of previous and current implementation should also go there? | 15:56 |
lambday | for linear time mmd and quadratic time mmd I mean | 15:57 |
lambday | (oh but in that case, we *have* to have both the implementation in the same branch which we don't have currently) | 15:57 |
lambday | HeikoS: btw dino also suggested that we put kernel selection stuff under a base CKernelSelection instead of CMMDKernelSelection.. | 15:58 |
lambday | so inside feature selection stuff, we can use that instead | 15:58 |
lambday | which I think is a great idea for future kernel selection stuffs we might have | 15:59 |
-!- thoralf [~thoralf@91-66-177-150-dynip.superkabel.de] has joined #shogun | 16:22 | |
thoralf | Heyho. | 16:22 |
-!- jiaolong [9e6d09e4@gateway/web/freenode/ip.158.109.9.228] has joined #shogun | 16:54 | |
-!- pickle27 [~pickle27@130.15.32.56] has joined #shogun | 16:55 | |
-!- pickle27_ [~pickle27@130.15.32.56] has joined #shogun | 16:56 | |
-!- pickle27 [~pickle27@130.15.32.56] has quit [Read error: Connection reset by peer] | 16:56 | |
-!- fabian__ [8184de63@gateway/web/freenode/ip.129.132.222.99] has joined #shogun | 16:57 | |
-!- pickle27_ [~pickle27@130.15.32.56] has quit [Quit: This computer has gone to sleep] | 17:31 | |
@wiking | HeikoS: ping | 17:34 |
-!- nystisil [~qdrgsm@80.252.20.67] has quit [Ping timeout: 272 seconds] | 17:38 | |
-!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.] | 17:41 | |
-!- khalednasr [~k.nasr92@41.69.211.87] has quit [Quit: Leaving] | 17:42 | |
-!- pickle27 [~pickle27@130.15.32.56] has joined #shogun | 17:46 | |
-!- zxtx__ [~zv@g225019108.adsl.alicedsl.de] has joined #shogun | 18:09 | |
-!- jiaolong [9e6d09e4@gateway/web/freenode/ip.158.109.9.228] has quit [] | 18:12 | |
-!- zxtx_ [~zv@f052018195.adsl.alicedsl.de] has quit [Ping timeout: 255 seconds] | 18:12 | |
-!- weqewewqew [~visof@unaffiliated/visof] has quit [Ping timeout: 272 seconds] | 18:23 | |
-!- pickle27 [~pickle27@130.15.32.56] has quit [Quit: This computer has gone to sleep] | 18:30 | |
-!- thoralf [~thoralf@91-66-177-150-dynip.superkabel.de] has quit [Quit: Konversation terminated!] | 18:32 | |
-!- abinash_panda [~abinash@117.214.89.52] has joined #shogun | 18:33 | |
-!- lambday [67157f4c@gateway/web/freenode/ip.103.21.127.76] has quit [Ping timeout: 246 seconds] | 18:47 | |
@HeikoS | wiking: po | 18:52 |
@HeikoS | wiking: ng | 18:53 |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 19:29 | |
shogun-notifier- | shogun: lambday :feature/mmd * f89a274 / / (6 files): https://github.com/shogun-toolbox/shogun/commit/f89a274ca6a0726baed5c4d49c2e5c2f1e8b95a3 | 19:29 |
shogun-notifier- | shogun: B-test MMD added | 19:29 |
shogun-notifier- | shogun: Heiko Strathmann :feature/mmd * 15f8032 / / (6 files): https://github.com/shogun-toolbox/shogun/commit/15f80322801a2f863009a9800c24c8d3354e26fb | 19:29 |
shogun-notifier- | shogun: Merge pull request #2317 from lambday/feature/btest | 19:29 |
shogun-notifier- | shogun: | 19:29 |
shogun-notifier- | shogun: B-test MMD added | 19:29 |
-!- kislay [~Abhijeet@103.240.205.138] has quit [Ping timeout: 264 seconds] | 19:31 | |
-!- thoralf [~thoralf@91-65-141-60-dynip.superkabel.de] has joined #shogun | 19:40 | |
thoralf | Heyho. | 19:40 |
@HeikoS | thoralf: hey! | 19:41 |
-!- khalednasr [~k.nasr92@41.69.211.87] has joined #shogun | 20:15 | |
-!- pickle27 [~pickle27@130.15.32.56] has joined #shogun | 20:25 | |
-!- pickle27 [~pickle27@130.15.32.56] has quit [Quit: This computer has gone to sleep] | 20:56 | |
abinash_panda | thoralf: hey ! | 20:59 |
-!- HeikoS [~heiko@nat-240-155.internal.eduroam.ucl.ac.uk] has quit [Quit: Leaving.] | 21:00 | |
-!- lisitsyn [~qdrgsm@80.252.20.67] has joined #shogun | 21:10 | |
lisitsyn | khalednasr: hey what's up | 21:10 |
khalednasr | lisitsyn, hey, starting to work on deep belief nets | 21:11 |
lisitsyn | khalednasr: could you please explain your last PR :D | 21:11 |
lisitsyn | khalednasr: I merged it as it is totally ok code-wise :D | 21:11 |
lisitsyn | but I don't really know what's this sampling with evidence | 21:12 |
khalednasr | lisitsyn, ahh, well normally after you train an rbm on some data you'll want to draw samples from the marginal P(data) to see what the rbm thinks the data looks like | 21:13 |
khalednasr | thats what the regular sample() functions do | 21:13 |
lisitsyn | khalednasr: will you see some different blobs for different classes? | 21:13 |
khalednasr | but if your data is divided into two groups, say pixels and labels | 21:13 |
lisitsyn | ahh so this is where evidence comes from | 21:14 |
khalednasr | by sampling with evidence I mean sampling from P(pixels | labels) | 21:14 |
lisitsyn | cool I see | 21:15 |
lisitsyn | khalednasr: this marginalized P(data) | 21:16 |
lisitsyn | how does it look? | 21:16 |
lisitsyn | what are the samples? | 21:16 |
khalednasr | lisitsyn, if you're training on digits, the samples will look like digits | 21:16 |
lisitsyn | ahh got it | 21:16 |
lisitsyn | khalednasr: how does it work? :D | 21:17 |
lisitsyn | khalednasr: I guess that's possible for RBMs but not for any other kind of NN, right? | 21:17 |
khalednasr | lisitsyn, it's like fitting a multivariate gaussian to some data, after you learn the mean and the variance, you can draw samples from the distribution and they will look somewhat like the data you trained on | 21:18 |
lisitsyn | got it | 21:18 |
lisitsyn | cool | 21:18 |
lisitsyn | thanks for this explanation! | 21:19 |
khalednasr | lisitsyn, for rbms, you learn the weights and then use gibbs sampling starting from some random states and you end up with samples from the model | 21:19 |
khalednasr | it only works for rbms | 21:19 |
khalednasr | you're welcome :) | 21:19 |
-!- abinash_panda [~abinash@117.214.89.52] has quit [Ping timeout: 255 seconds] | 21:20 | |
lisitsyn | khalednasr: wow you're just on time with your schedule | 21:24 |
lisitsyn | that's scary | 21:24 |
lisitsyn | :D | 21:24 |
khalednasr | lisitsyn, except for the GPU stuff :D | 21:24 |
lisitsyn | that's ok | 21:24 |
lisitsyn | we have to push it though I guess | 21:25 |
lisitsyn | but you finished RBMs exactly the date you planned :D | 21:25 |
lisitsyn | that's scary! | 21:25 |
khalednasr | lisitsyn, haha :D | 21:25 |
-!- abinash_panda [~abinash@117.214.89.52] has joined #shogun | 21:46 | |
-!- khalednasr [~k.nasr92@41.69.211.87] has quit [Ping timeout: 255 seconds] | 22:04 | |
-!- Netsplit *.net <-> *.split quits: sonne|work, lisitsyn, naywhayare, zxtx, soumyaC | 22:19 | |
-!- Netsplit over, joins: lisitsyn, soumyaC, zxtx, naywhayare, sonne|work | 22:20 | |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 22:29 | |
-!- pickle27 [~pickle27@24.114.81.98] has joined #shogun | 22:46 | |
-!- pickle27 [~pickle27@24.114.81.98] has quit [Ping timeout: 255 seconds] | 22:52 | |
-!- khalednasr [~k.nasr92@41.69.225.75] has joined #shogun | 23:43 | |
--- Log closed Wed Jun 18 00:00:37 2014 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!