Re: Machine learning support

[Alois Schloegl]

On 2017-12-28 16:08, Juan Pablo Carbajal wrote:
> Hi,
> 
> I work on ML mainly using Python (scikitlearn, mltk, theano, etc) and
> Octave (gpml, stk, optim, etc). I stay away from R when possible. I
> also use C/C++ (Shogun, MLPack, etc...) which many times can provide
> an oct file interface.
> 
> ML as a topic is too wide. You rather focus on one aspect of ML and
> write your package, examples are nnet, fuzzy, gpml.
> If your plan is to do ML on text, I would say that Octave is not the
> best choice, in this cas eI woudl stick to Python.
> 
> Take a look at Julia language as well, there is plenty of packages
> already for that one.
> 
> Cheers
> 
> _______________________________________________
> Help-octave mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/help-octave
> 


Hi Juan,


I recommend taking a look at the "NaN-toolbox - A statistics and machine
learning toolbox ..." [1].

It contains a number of machine learning methods. For more details see
the help functions of train_sc, test_sc, classify, and xval:

  https://octave.sourceforge.io/nan/function/train_sc.html
  https://octave.sourceforge.io/nan/function/test_sc.html
  https://octave.sourceforge.io/nan/function/classify.html
  https://octave.sourceforge.io/nan/function/xval.html


Cheers,
  Alois



[1] The NaN-toolbox: A statistics and machine learning toolbox for
Octave and Matlab® for data with and w/o MISSING VALUES encoded as NaN's.       

https://pub.ist.ac.at/~schloegl/matlab/NaN/
https://octave.sourceforge.io/nan/index.html
https://mloss.org/software/view/206/