--- Log opened Wed Sep 26 00:00:17 2012 | ||
-!- hoijui [~hoijui@dslb-088-074-107-101.pools.arcor-ip.net] has quit [Ping timeout: 248 seconds] | 00:36 | |
-!- heiko1 [~heiko@host86-181-80-213.range86-181.btcentralplus.com] has quit [Quit: Leaving.] | 01:16 | |
shogun-buildbot | build #114 of nightly_default is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_default/builds/114 | 03:47 |
---|---|---|
-!- zxtx [~zv@cpe-75-83-151-252.socal.res.rr.com] has left #shogun ["Leaving"] | 03:59 | |
-!- hoijui [~hoijui@dslb-088-074-107-101.pools.arcor-ip.net] has joined #shogun | 07:28 | |
-!- blackburn [5bdfb203@gateway/web/freenode/ip.91.223.178.3] has joined #shogun | 09:11 | |
wiking | blackburn: yo | 09:22 |
blackburn | wiking: hey | 09:22 |
blackburn | wiking: ha! I bet these slep leaks blocks you from sleeping :) | 09:47 |
wiking | blackburn: hhehehe no not really just dont want to start with other latent stuff | 09:48 |
wiking | :) | 09:48 |
blackburn | thanks, I merged that | 09:49 |
wiking | so i've started up fixing the static analyzer stuff | 09:49 |
shogun-buildbot | build #545 of deb3 - modular_interfaces is complete: Failure [failed test lua_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/545 blamelist: Viktor Gal <viktor.gal@maeth.com> | 10:16 |
wiking | heheh wazntme :) | 10:21 |
blackburn | wiking: yes, I didn't manage to fix up lua yet | 10:23 |
shogun-buildbot | build #546 of deb3 - modular_interfaces is complete: Failure [failed test lua_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/546 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 10:37 |
sonne|work | blackburn: the build braker is in da hous | 10:38 |
sonne|work | e | 10:38 |
blackburn | sonne|work: lua is undebuggable | 10:38 |
sonne|work | I see | 10:39 |
sonne|work | This thing sporadically pops up on the bb | 10:39 |
sonne|work | but never locally... | 10:39 |
blackburn | sonne|work: no, it actually hangs on my machine | 10:41 |
sonne|work | but then valrind / gdb and voila! | 10:43 |
blackburn | sonne|work: I've got recursive calls of typemaps | 10:44 |
blackburn | neither gdb nor valgrind pukes :D | 10:44 |
blackburn | it is just eats all my 8gb of memory and dies | 10:45 |
blackburn | it just* | 10:45 |
sonne|work | well ok then check the typemaps :) | 10:46 |
blackburn | yeah I did (twice!!!!!) :) | 10:49 |
blackburn | well, I'll get back to that soon | 10:49 |
-!- heiko [~heiko@host86-182-167-91.range86-182.btcentralplus.com] has joined #shogun | 11:18 | |
blackburn | http://en.wikipedia.org/wiki/Troxler%27s_fading | 11:59 |
heiko | blackburn, wiking, I solved all problems in another way :) | 13:15 |
heiko | blackburn, no streaming kernel needed, I just extended streaming features and use normal kernel | 13:16 |
heiko | wiking, I have removed this method I mentioned. Its possible to let the caller decide at another point in the code | 13:16 |
shogun-buildbot | build #438 of deb2 - static_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/438 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com> | 13:25 |
heiko | dont worry buildbot, the last build of the 50 works :) | 13:26 |
blackburn | heiko: it doesn't, lua is still broken :) | 13:31 |
heiko | blackburn, lol :) | 13:31 |
-!- hoijui [~hoijui@dslb-088-074-107-101.pools.arcor-ip.net] has quit [Ping timeout: 240 seconds] | 13:35 | |
wiking | heiko: :) | 13:36 |
heiko | this streaming two-sample test is really cool now: 500 millions samples and more | 13:36 |
heiko | gotta go, see you later guys | 13:36 |
blackburn | 500 millions?? | 13:36 |
blackburn | how? | 13:36 |
heiko | blackburn, yeah :) | 13:36 |
blackburn | why? :D | 13:36 |
heiko | well streaming | 13:37 |
heiko | some problems are so hard that they need this many samples | 13:37 |
blackburn | how does that work? | 13:37 |
blackburn | I thought you create NxN matrix | 13:37 |
heiko | streaming data, so does not have to be stored in mem | 13:37 |
heiko | no | 13:37 |
blackburn | 2 examples by step? | 13:37 |
heiko | just need trace of kernel matrix | 13:37 |
blackburn | 2x2? | 13:37 |
blackburn | ah | 13:37 |
heiko | kind of yeah | 13:37 |
blackburn | interesting | 13:37 |
blackburn | remarkable thing to mention | 13:38 |
blackburn | like | 13:38 |
blackburn | hey guys we are working with data with 500 millions examples | 13:38 |
blackburn | what's up with your dlib/scikits/weka? | 13:38 |
blackburn | :D | 13:38 |
heiko | indeed, thats what I did in my thesis ;) | 13:38 |
heiko | more interesting is: there are problems that *need* that much data | 13:38 |
heiko | otherwise you cannot solve them since they are too hard | 13:38 |
blackburn | what is the problem? | 13:39 |
wiking | lol | 13:39 |
heiko | http://herrstrathmann.de/images/blobs.png | 13:39 |
blackburn | that's like such a big data one can estimate probability distribution pretty precisely | 13:39 |
heiko | to distinguish between the red and the blue | 13:39 |
blackburn | is that possible? | 13:39 |
heiko | (with other parameters than in the image, but thats the principle) | 13:39 |
wiking | heiko: and what machine do u use with the kernel? | 13:39 |
heiko | CLinearTimeMMD, its not a machine | 13:40 |
heiko | statistical test | 13:40 |
wiking | ah yeah | 13:40 |
heiko | but I mean two-sample testing and binary classification is similar | 13:40 |
heiko | nature | 13:40 |
blackburn | so H0 is like blue | 13:40 |
blackburn | and H1 is red | 13:40 |
blackburn | ? | 13:40 |
wiking | heiko: so u reckon it should work with a binary machine as well? | 13:40 |
wiking | *binary classifier... | 13:40 |
heiko | blackburn, yes | 13:41 |
heiko | no | 13:41 |
heiko | ehm, no | 13:41 |
heiko | H0: p=q | 13:41 |
heiko | H1: p!=q | 13:41 |
heiko | where p is read and q is blue | 13:41 |
heiko | wiking, well I dont know a machine that can handle that much data | 13:41 |
heiko | we have online lib-linear right? | 13:42 |
wiking | heiko: apart from online | 13:42 |
wiking | or like pegasus-svm | 13:42 |
heiko | ah ok | 13:42 |
wiking | so yeah it would only work with that ... | 13:42 |
heiko | year, easily classifiable data is also easy to distinguish in a two-sample test | 13:42 |
heiko | the dataset is quite interesting | 13:42 |
heiko | since the differences are hidden inside, at a different scalling that the overall data has | 13:43 |
blackburn | heiko: so we only say it is the same distribution or it is not | 13:43 |
heiko | blackburn, yes, but one tries to find a difference, that is "pull the distributions apart", so quite similar to classification | 13:43 |
heiko | when you can pull them apart, you can put a hyperplane in between | 13:43 |
blackburn | I see | 13:43 |
heiko | I gotta go now, see you later :) | 13:44 |
wiking | laterz | 13:44 |
shogun-buildbot | build #547 of deb3 - modular_interfaces is complete: Failure [failed test lua_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/547 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com> | 13:44 |
-!- heiko [~heiko@host86-182-167-91.range86-182.btcentralplus.com] has quit [Quit: Leaving.] | 13:44 | |
shogun-buildbot | build #439 of deb2 - static_interfaces is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/439 | 13:51 |
shogun-buildbot | build #548 of deb3 - modular_interfaces is complete: Failure [failed test lua_modular] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/548 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com> | 13:59 |
-!- heiko [~heiko@host86-182-167-91.range86-182.btcentralplus.com] has joined #shogun | 14:27 | |
-!- heiko [~heiko@host86-182-167-91.range86-182.btcentralplus.com] has quit [Quit: Leaving.] | 16:38 | |
-!- blackburn [5bdfb203@gateway/web/freenode/ip.91.223.178.3] has quit [Quit: Page closed] | 17:16 | |
-!- hoijui [~hoijui@dslb-088-074-107-101.pools.arcor-ip.net] has joined #shogun | 17:46 | |
-!- in3xes [~in3xes@122.174.80.141] has joined #shogun | 18:00 | |
* wiking bangs the head into the wall... found a bug! :D | 18:35 | |
-!- blackburn [~blackburn@188.168.14.95] has joined #shogun | 18:57 | |
-!- romi__ [~mizobe@187.57.9.8] has quit [Ping timeout: 246 seconds] | 19:36 | |
-!- romi__ [~mizobe@187.101.115.113] has joined #shogun | 19:50 | |
shogun-buildbot | build #440 of deb2 - static_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb2%20-%20static_interfaces/builds/440 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 19:57 |
shogun-buildbot | build #434 of bsd1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/bsd1%20-%20libshogun/builds/434 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 19:58 |
-!- in3xes_ [~in3xes@122.174.90.177] has joined #shogun | 19:58 | |
-!- in3xes [~in3xes@122.174.80.141] has quit [Ping timeout: 265 seconds] | 20:01 | |
wiking | yes !!! SO-BMRM: 100.00 | 20:12 |
wiking | \o/ | 20:12 |
wiking | so it actually worx !!! | 20:14 |
shogun-buildbot | build #31 of rpm1 - libshogun is complete: Failure [failed test] Build details are at http://www.shogun-toolbox.org/buildbot/builders/rpm1%20-%20libshogun/builds/31 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 20:16 |
shogun-buildbot | build #549 of deb3 - modular_interfaces is complete: Failure [failed test libshogun] Build details are at http://www.shogun-toolbox.org/buildbot/builders/deb3%20-%20modular_interfaces/builds/549 blamelist: Sergey Lisitsyn <lisitsyn.s.o@gmail.com> | 20:26 |
blackburn | wiking: great | 20:33 |
wiking | blackburn: yep in the meanwhile i even invented a new way of training the latent so-svm ;) | 20:42 |
wiking | to have a better convergence | 20:42 |
blackburn | without a train set? | 20:42 |
wiking | i have a train set | 20:42 |
wiking | :P | 20:42 |
blackburn | bad | 20:42 |
blackburn | I'd like a method without train set | 20:42 |
blackburn | :) | 20:42 |
wiking | nocando | 20:43 |
wiking | :) | 20:43 |
blackburn | wiking: wanna try to beat guys on digit recognition? http://www.kaggle.com/c/digit-recognizer | 20:44 |
blackburn | kind of test | 20:45 |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has joined #shogun | 21:04 | |
wiking | blackburn: kagggle and no prize? :) | 21:12 |
blackburn | wiking: I do not expect any prize to receive without experience.. | 21:12 |
blackburn | btw | 21:13 |
blackburn | check that | 21:14 |
blackburn | https://gist.github.com/3789950 | 21:14 |
wiking | what's the time? :) | 21:14 |
blackburn | looks like useful thing for ml, I am now thinking about a way to register local context variables and save them | 21:14 |
blackburn | time.time? | 21:15 |
blackburn | well, time | 21:15 |
blackburn | :) | 21:15 |
blackburn | with thing looks like pretty compact way to designate scopes where you need to save something or measure | 21:15 |
wiking | http://worrydream.com/LearnableProgramming/ | 21:54 |
wiking | hehe | 21:54 |
blackburn | tl;dr | 21:57 |
blackburn | ?? | 21:57 |
blackburn | oops :D | 21:57 |
wiking | :> | 21:58 |
-!- n4nd0 [~nando@s83-179-44-135.cust.tele2.se] has quit [Quit: Lost terminal] | 22:40 | |
-!- blackburn [~blackburn@188.168.14.95] has quit [Quit: Leaving.] | 23:44 | |
-!- hoijui [~hoijui@dslb-088-074-107-101.pools.arcor-ip.net] has quit [Ping timeout: 248 seconds] | 23:48 | |
--- Log closed Thu Sep 27 00:00:17 2012 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!