[Scilab-Dev] [SOCIS]Add advanced linear programming capabilities under Scilab.

[amandawy.chen]

Hi all,

I'm Wanying Chen from France, interested in adding advanced LP optimization
extensions in Scilab during SOCIS 2014.
I have proficient C++ programming experience and well prepared for this
project.

Current Scilab could not easily used for solving optimization models in
machine learning such as SVM and non-differentiable models such as L1-norm
term in Lasso.

So, I would like to add the OptiML and DFO packages into Scilab. Also, since
the conventional gradient descent method has slow convergence, the SMI for
stochastic gradient descent should also be added.

In my mind, I'd like to get these packages integrated into Scilab and
enhance the current documentation about optimization especially the problems
related to machine learning.
Is it good for a SOCIS project? Where should I start with?
Looking forward to your advise.


Best,
Wanying



--
View this message in context: http://mailinglists.scilab.org/SOCIS-Add-advanced-linear-programming-capabilities-under-Scilab-tp4030542.html
Sent from the Scilab developers - Mailing Lists Archives mailing list archive at Nabble.com.