What is the rate of convergence of the hebbian algorithm? It appears to be another incarnation of a classical algorithm used by numerical linear algebraist.
On 18 June 2015 at 15:34, Kyle Kastner <kastnerk...@gmail.com> wrote: > Another citation for hebbian approach - it is related to this > http://onlinelibrary.wiley.com/doi/10.1207/s15516709cog0901_5/pdf > > On Thu, Jun 18, 2015 at 10:25 AM, Kyle Kastner <kastnerk...@gmail.com> > wrote: > >> Yes agreed - though I would also guess the intermediate memory blowup >> could help speed, though I haven't tested. I guess it comes back to the >> question, has anyone done MiniBatchKMeans on MNIST? To be honest I don't >> recall the original the question from ~1 year ago, but would be very >> surprised if there was a problem. I use MiniBatchKMeans (and this hebbian >> thing) on *much* larger datasets... >> >> On Thu, Jun 18, 2015 at 10:17 AM, Andreas Mueller <t3k...@gmail.com> >> wrote: >> >>> You could implement Lloyds algorithm in as little code, too. >>> One of the reasons that the sklearn implementation is much longer is >>> that it doesn't do fancy indexing and avoids large intermediate arrays. >>> >>> >>> >>> On 06/18/2015 10:09 AM, Kyle Kastner wrote: >>> >>> I don't know if it is faster or better - but the learning rule is >>> insanely simple and it is hard to believe there could be *anything* much >>> faster. It is ten lines - won't copy it here cause the license is longer >>> than the implementation! >>> >>> Given the connection between PCA and K-means, this implementation >>> (Matlab...) is also related >>> http://homepages.cae.wisc.edu/~ece539/matlab/ghafun.m >>> >>> This points to this paper: >>> http://courses.cs.washington.edu/courses/cse528/09sp/sanger_pca_nn.pdf >>> >>> Basically this is the neural net approach to K-means. I have asked if >>> there is a paper ref - though it might be "too easy" to have a real paper. >>> >>> >>> On Thu, Jun 18, 2015 at 9:58 AM, Andreas Mueller <t3k...@gmail.com> >>> wrote: >>> >>>> >>>> >>>> On 06/18/2015 09:48 AM, Kyle Kastner wrote: >>>> > This link should work http://www.cs.toronto.edu/~rfm/code.html >>>> > <http://www.cs.toronto.edu/%7Erfm/code.html> >>>> Is that faster / better than minibatch k-means? Is there a paper? >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> _______________________________________________ >>>> Scikit-learn-general mailing list >>>> Scikit-learn-general@lists.sourceforge.net >>>> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general >>>> >>> >>> >>> >>> ------------------------------------------------------------------------------ >>> >>> >>> >>> _______________________________________________ >>> Scikit-learn-general mailing >>> listScikit-learn-general@lists.sourceforge.nethttps://lists.sourceforge.net/lists/listinfo/scikit-learn-general >>> >>> >>> >>> >>> ------------------------------------------------------------------------------ >>> >>> _______________________________________________ >>> Scikit-learn-general mailing list >>> Scikit-learn-general@lists.sourceforge.net >>> https://lists.sourceforge.net/lists/listinfo/scikit-learn-general >>> >>> >> > > > ------------------------------------------------------------------------------ > > _______________________________________________ > Scikit-learn-general mailing list > Scikit-learn-general@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/scikit-learn-general > >
------------------------------------------------------------------------------
_______________________________________________ Scikit-learn-general mailing list Scikit-learn-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
