[ml] Linux image for ML Course?

[Nitin Borwankar]

The class code works perfectly with Octave.  MATLAB is not needed.  I had
never coded in Octave before - the tutorial lessons in the class are
sufficient. The learning curve is not very steep and I dont consider myself
an expert programmer.  Andrew Ng is an excellent teacher.  If you haven't
taken an ML class yet - I would recommend the Coursera class first.  Unless
you are hungry for the math in which case take the Caltech class.

Nitin


------------------------------------------------------------------
Nitin Borwankar
nborwankar at gmail.com


On Sat, Apr 6, 2013 at 1:46 PM, John Ellis <neurofog at gmail.com> wrote:

> The MATLAB requirement is almost a show-stopper. With some "skill" it may
> be possible to convert over to GNU Octave or similar, but Python ML/Math
> libraries are more accessible and easier learning curve than MATLAB/Octave
> would be.
>
> As I recall last time, Jared created a VirtualBox image with Linux &
> Python libraries preinstalled. If there is enough interest, I could do the
> same for this round.
>
> Cheers!
> -John
>
> On Fri, Apr 5, 2013 at 10:33 PM, Nitin Borwankar <nborwankar at gmail.com>wrote:
>
>> I am also interested in a study group for the Caltech class.
>>
>> The Stanford one taught at Coursera was very "application" oriented.  The
>> Caltech one is strong on theory but also has some practical aspects.
>> Probably more like the Stanford one taught at Stanford. IIRC there are two
>> versions the Coursera one has an "a" suffix which means it's focused on
>> "applications" not theory.
>>
>> Also the Stanford one required MATLAB for problems.  The Caltech one lets
>> you use whatever language you want.
>>
>> Nitin
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.noisebridge.net/pipermail/ml/attachments/20130406/b46621b9/attachment-0001.html>