Hi James, Thank you, and you are absolutely right, now I have minimized my proposal to focus on implementing deep belief nets and auto encoders, and made them subjective to change, as there might be more suitable DL algorithms to implement as a starter for scikit-learn.
In case you would like to follow, here is the updated proposal, https://google-melange.appspot.com/gsoc/proposal/review/google/gsoc2013/issamou/1# Thank you, --Issam On 5/3/2013 4:38 PM, James Bergstra wrote: > "A flag into your hands" as they say, Issam. But do bear in mind that > many of those algorithms that you've budgeted only a week to > implement, may well take a week to *run*, on the kinds of data sets > for which they have been tested in the literature. Tuning > hyperparameters of such costly algorithms takes a lot of compute > resources, which become expensive at that scale. In my humble opinion, > I believe your proposal is more appropriate for about 3 PhDs, or > perhaps a small dev team over a few years, than a google summer of > code project. If you could get any *one* of those algorithms into a > state worthy of consideration by sklearn it would be a great > achievement that has so far eluded many skilled people. If you don't > adjust your proposal to be more modest, you run a risk of appearing > like you haven't really thought things through. > > - James > > ------------------------------------------------------------------------------ Get 100% visibility into Java/.NET code with AppDynamics Lite It's a free troubleshooting tool designed for production Get down to code-level detail for bottlenecks, with <2% overhead. Download for free and get started troubleshooting in minutes. http://p.sf.net/sfu/appdyn_d2d_ap2 _______________________________________________ Scikit-learn-general mailing list Scikit-learn-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/scikit-learn-general