Apologizes Matthew, the comment about Windows relates to my experience a few months ago when initially setting up NuPIC. I'm more familiar with Perforce and Visual C++ & C#. I think the confusing thing at that time was the ongoing? split of the core. Trying to define a clean way of keeping NuPIC up to date. It's always been quicker for me to use C#, or C++, for prototyping. My Python experience is limited, so as such various stumbling through.
I'm back from vacation on Sunday, so will look next week at getting critic up and running and determine whether just the core is needed, how well cmake makes VC projects? Basically attack critic and NuPIC within a Windows environment and whether cygwin, python, etc. is needed. Just stepping in.. The genre classification seems to me to be a nice way to jump in. I came across the dataset through Yann LeCun's recent presentations (long, but some quite a few chuckles); Deep Learning: The Theoretician's Nightmare or Paradise? https://www.youtube.com/watch?v=HrMU1GgyxL8 This lead to Bob Sturm et al work on the dataset, and in particular this paper; On Automatic Music Genre Recognition by Sparse Representation Classification using Auditory Temporal Modulations http://vbn.aau.dk/files/60313218/genreSturmNoorzad20120116.pdf Regards, Richard On Fri, Aug 8, 2014 at 7:16 PM, Matthew Taylor <[email protected]> wrote: > Richard, that sounds interesting. I think it would be possible. We > have an intern experimenting with nupic.critic doing something > similar. I'm not sure what you mean about getting NuPIC running under > Windows, that seems like a completely different problem. Can you > explain? > --------- > Matt Taylor > OS Community Flag-Bearer > Numenta > > > On Fri, Aug 8, 2014 at 12:40 AM, Richard Crowder <[email protected]> wrote: > > Hi Matt, > > > > I'm wondering whether it is worth expanding critic to handle genre > > classification? Modifying the current code to allow for a dataset like > > GTZAN (modified through Bob L. Sturm's work) to setup an initial > > classifier. For me it sounds like a great way to get NuPIC running > > under Windows (I've installed and run tests via VMs and cygwin) and > > learning the internals of NuPIC. Does that sounds like a good idea to > > pursue? > > > > Regards, Richard. > > > > On Sat, Jul 12, 2014 at 4:59 PM, Matthew Taylor <[email protected]> > wrote: > >> On Fri, Jul 11, 2014 at 11:04 PM, Alexander Hirner <[email protected]> > >> wrote: > >>> > >>> Great test case. As far as I understood, you build separate models for > >>> each frequency bin and aggregate a total anomaly score by an arbitrary > >>> number of how many single anomaly score must be above a certain > treshold. > >> > >> > >> Correct. And the # of anomalyLikelihood values that cause the trigger is > >> configurable, as well as the anomalyLikelihood threshold itself. You > can do > >> a lot of tuning with these two values, which I haven't done much of > yet. I'm > >> still trying to find the best parameters for these for the music I'm > passing > >> in. > >> > >>> > >>> What do you think about dumping all the bins into one TP at once? In > that > >>> way, the model could also capture covariance between the frequency > bins and > >>> derive on a conclusion of an aggregate anomaly score on its own. > >> > >> > >> I wouldn't know how to do that. The input bin values would need to be > >> converted into SDRs, which is what the SP does. The only time I've ever > fed > >> data directly into the TP is when my data was already in SDR format > (from > >> Cortical.IO API). > >> > >> --------- > >> Matt Taylor > >> OS Community Flag-Bearer > >> Numenta > >> > >> _______________________________________________ > >> nupic mailing list > >> [email protected] > >> http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org > >> > > > > _______________________________________________ > > nupic mailing list > > [email protected] > > http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org > > _______________________________________________ > nupic mailing list > [email protected] > http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org >
_______________________________________________ nupic mailing list [email protected] http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
