On May 5, 12:42 am, Sergey Bochkanov <[email protected]>
wrote:
> Hello, William.
>
> You wrote 4 мая 2010 г., 22:34:58:
>
> > Could you write a little more to sage-devel about why mathematicians
> > might care about "support vector machines" -- it's possible that most
> > people reading this have never heard of them.
>
> Being  sage-devel  reader  for  a  while I may say that SAGE users are
> mostly  interested  in  symbolic algebra, graph theory, other kinds of
> "exact algebra" (finite rings, etc.). Interest in numerical processing
> is much more lower.

You're right. I guess that's the purpose of this thread -- to gauge
the interest of the Sage community in routines for using SVMs. Just
because the Sage community in general isn't interested in a component,
however, doesn't mean that it couldn't be useful to some people.

>
> Looks like we (Open Source Community) have:
> * R - for statistical processing
> * Octave - for numerical processing
> * SAGE - for symbolic/"exact" problems
>
> What do you think about such specialization? Is it intentional?

I'm sure its intentional. Its good for a project to have both scope
and vision. However, there are circumstances in which you may want to
expand a project's scope. And SVM-Light doesn't have to be an addition
to Sage's core -- it could be an optional SPKG that interested users
may install at their leisure.

> P.S.  I've thought on writing SAGEable Python wrapper for ALGLIB. This
> idea  came to me when I worked with automatically generated C# wrapper
> for  MPIR.  It started as MPIR-related project, but now I see that the
> same technology can be applied to other software projects.
>
> I  know  that  SAGE  already  has  GSL wrapper, but two projects don't
> overlap  each other. Some functionality is present only in GSL, some -
> only in ALGLIB. But is numerical analysis really needed in the SAGE?
>
> --
> With best regards,
>  Sergey                          mailto:[email protected]

In my opinion SVM functionality could be a useful addition to Sage.
For ML researchers, being able to execute a support vector machine and
then graph and manipulate the results all within an interactive
environment could be very useful. However, if no one is interested
then I won't bother expending the effort!

-- Bill

-- 
To post to this group, send an email to [email protected]
To unsubscribe from this group, send an email to 
[email protected]
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org

Reply via email to