[ml] what should we do next? deep nets, reinforcement learning, complex systems?

[Brian Morris]

I'm for starting with the Deep Nets first.

For the RL stuff, I am thinking of it for myself as a research project (ie
longish term).  So far I have read one chapter in a book on ML and I have
been collecting papers and code packages and examples from the Web, and I am
in the process of sorting through / evalutating possibilities. I'd like to
talk or correspond with Karim, I like him leaving open both possibility of
hearing / involvement.

Part of the reason I am interested in RL is it brings together a lot of
things, hopefully in sort of a manageable, starting out kind of way.
Bringing in dynamics and simulation with a single agent, rather than a
system. As far as programming goes, a chance to develop Python skills with
given supporting packages for. As far as application goes possibly
connecting to microcontrollers as in feedback / control systems which is
interest of electronics group at noisebridge (although this would require
lower level C programming )

Brian


On Thu, Jul 28, 2011 at 1:32 PM, Mike Schachter <mike at mindmech.com> wrote:

> Awesome - it would be great if anyone with any interest gives their
> vote as well!
>
>
>
> On Thu, Jul 28, 2011 at 1:31 PM, Karim Wahba <karim.a.wahba at gmail.com>
> wrote:
> > hi All,
> >
> > thanks for organizing all this. If we're taking votes, I'd be interested
> in
> > hearing more/being involved in the reinforcement learning talks.
> >
> > thx,
> > Karim
> >
> > On Thu, Jul 28, 2011 at 12:53 PM, Mike Schachter <mike at mindmech.com>
> wrote:
> >>
> >> Hey everyone,
> >>
> >> Let's put our heads together and decide what we're going
> >> to do for the next few Wednesdays. Probably no formal
> >> workshops, as they take some time, but it would be good to
> >> get together and have an informal but focused discussion
> >> about a specific topic.
> >>
> >> On the agenda are deep nets, reinforcement learning, probably
> >> with markov decision processes, and complex systems.
> >>
> >> Kai - are you interested in putting together an informal meetup
> >> about complex systems? If so, when would be a good day for
> >> you?
> >>
> >> Brian - what/when do you think you'll be comfortable guiding a
> >> formal meetup on reinforcement learning and markov stuff? I can
> >> help out with the markov part.
> >>
> >> Ron - I'm interested in doing informal deep net stuff either next
> >> week or the week after, but let's see how this thread pans out.
> >>
> >>  mike
> >> _______________________________________________
> >> ml mailing list
> >> ml at lists.noisebridge.net
> >> https://www.noisebridge.net/mailman/listinfo/ml
> >
> >
> _______________________________________________
> ml mailing list
> ml at lists.noisebridge.net
> https://www.noisebridge.net/mailman/listinfo/ml
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.noisebridge.net/pipermail/ml/attachments/20110728/1bc5f7b2/attachment-0003.html>