[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: new task: machine-learning?

On Thu, 10 Jun 2010, Soeren Sonnenburg wrote:
> > I wanted to check with you first, either it looks reasonable/desired
> > before I commit it.  Please find a tentative task file attached.
> where do you draw the line that something is machine learning? I mean
> many ML people use matlab (or its free variant octave) or python-numpy,
> optimization packages (glpk) etc?

Well, and they all use algebra, but it is still not a higher math,

so, matlab, octave, numpy, scipy -- generic computational
frameworks/libraries.  I think I can't tell that about shogun, but I can
say that it was geared toward machine learning.  Would I be wrong?

but your point has merits, that all those generic libraries might have
components which fall into realm of machine learning (polynomial fitting
etc within scipy), but usually they would have more evolved components
outside (e.g. additional toolkit for a $ in Matlab, or scikit-learn for
scipy) and those would become a part of this task.  May be generic
frameworks (besides Matlab) could be suggested (weak dependency) as well
-- feel free to contribute.

> Soeren
=------------------------------   /v\  ----------------------------=
Keep in touch                    // \\     (yoh@|www.)onerussian.com
Yaroslav Halchenko              /(   )\               ICQ#: 60653192
                   Linux User    ^^-^^    [175555]

Reply to: