verne.freenode.net changed the topic of #mlpack to: http://www.mlpack.org/ -- We don't respond instantly... but we will respond. Give it a few minutes. Or hours. -- Channel logs: http://www.mlpack.org/irc/
benchmark has joined #mlpack
benchmark has quit [Client Quit]
vivekp has quit [Ping timeout: 264 seconds]
vivekp has joined #mlpack
vivekp has quit [Changing host]
vivekp has joined #mlpack
govg has quit [Ping timeout: 258 seconds]
mikeling has joined #mlpack
govg has joined #mlpack
chenzhe has quit [Ping timeout: 255 seconds]
govg has quit [Ping timeout: 240 seconds]
govg has joined #mlpack
govg has quit [Ping timeout: 240 seconds]
vivekp has quit [Ping timeout: 272 seconds]
vivekp has joined #mlpack
govg has joined #mlpack
govg has quit [Ping timeout: 260 seconds]
vivekp has quit [Ping timeout: 240 seconds]
vivekp has joined #mlpack
vivekp has quit [Ping timeout: 260 seconds]
vivekp has joined #mlpack
kris2 has joined #mlpack
kris2 has left #mlpack []
kris2 has joined #mlpack
govg has joined #mlpack
govg has quit [Ping timeout: 264 seconds]
govg has joined #mlpack
mentekid has quit [Ping timeout: 255 seconds]
govg has quit [Ping timeout: 255 seconds]
govg has joined #mlpack
kris2 has quit [Ping timeout: 255 seconds]
kris2 has joined #mlpack
mentekid has joined #mlpack
kris2 has quit [Ping timeout: 260 seconds]
kris2 has joined #mlpack
< kris2>
@lozhnikov you there
< kris2>
:lozhnikov
< lozhnikov>
kris2: hi
< kris2>
Hi
< kris2>
lozhnikov: just wanted to ask if you had gone through my proposal. Did you find the api okay.
< kris2>
Also should i first implement RBM
< kris2>
or GAN portion
< lozhnikov>
I think it doesn't matter
< lozhnikov>
As for me we have to correct the API
< kris2>
I think stacked GAN's are pretty difficult. That's why i wanted to start with RBM's
< kris2>
Ok, thats fine the API was just meant as a reference.
< lozhnikov>
ok. I am not sure of your variant of gibbs sampling
< kris2>
What exactly do you mean ?
< kris2>
Gibbs distribution is just sampling from a conditional distribution P(x_i| x_1,x_2,.....x_i-1,x_i+1,.....x_n) right?
< lozhnikov>
I mean the 10th page of your proposal
< lozhnikov>
actually we have to sample sample hidden variables from visible and then visible variables from hidden k times
< kris2>
Yes that's correct. Yes i see, my gibbs sampler implementation was more of a general gibbs sampler
< kris2>
I will change that implementation to more RBM specific
govg has quit [Ping timeout: 260 seconds]
< kris2>
I was also little unclear about block gibbs sampling
< kris2>
I will try to read about it tonight i think. We would need that for implementing the Persistence-CD
< lozhnikov>
I'll take a look at that.
< kris2>
:lozhnikov do you have any links/ books that you want to me study before we start coding.
< lozhnikov>
And I've got another question: It seems that your variant of RBM is not modular. As for me it is better to implement hidden and visible variables in terms of LayerType.
< lozhnikov>
As for me an implementation at deeplearning.net could be very helpful
< kris2>
Hmmm, okay i have gone through that, i will look at it again.
< kris2>
Will you be attending tomorrow meeting btw.
< lozhnikov>
yeah, I hope I'll be at home
< kris2>
cool, btw i know its a little late for a intro but still. i am masters student in iit hyderabad india. studying machine learning
< kris2>
my focus right now is submodular functions in ml. Earlier i was working on graph kernels
govg has joined #mlpack
< lozhnikov>
yeah, you pointed out that in your proposal:) did you work with neural networks before?
< kris2>
Yes, i have quite a lot actually.
< kris2>
mostly in the computer vision space. image captioning and stuff
< kris2>
though i really love reading Deep RL paper haven't gotten my hands dirty with RL yet.
< kris2>
What about you?
< lozhnikov>
I am a postgraduate student at the moscow state university. Right now I work on computer modelling of Navier-Stokes equations, gasdynamics equations and so on
< lozhnikov>
ML is my hobby
< kris2>
oh that's really cool.
govg has quit [Quit: leaving]
govg has joined #mlpack
mikeling has quit [Quit: Connection closed for inactivity]
chenzhe has joined #mlpack
kris2 has quit [Read error: Connection reset by peer]