[pymvpa] Which classifier to use?

[Francisco Pereira]

Hi Alyson,

Jo is right, you'd want an ordinal logistic regression (aka ordered logit).
Perhaps this implementation? (I'm afraid I haven't tried it...)

https://github.com/fabianp/minirank/blob/master/minirank/logistic.py

cheers,
Francisco

On Thu, Feb 18, 2016 at 9:42 AM, Jo Etzel <jetzel at wustl.edu> wrote:

> There's nothing at all wrong with a categorical classifier if you think of
> your four stimulus orientations as *categories* - four different stimulus
> types.
>
> But if you want to analyze the orientations as ordered classes (e.g., 15
> degrees half of 30 degrees; 75 more than 15), then yes, you will need a
> different classifier.
>
> Jo
>
>
>
>
> On 2/17/2016 4:58 PM, Alyson Saenz wrote:
>
>> Hey guys, I already analyzed my data but I think I can do better. I was
>> wondering if you can give me a hand.
>>
>>
>> I have a visual task with a continuous stimuli, oriented bars. I
>> presented several times specific degrees of the stimuli, so 15°, 30°, 60
>> and 75°. I tried to predict the orientation of the bars by analizing VI,
>> now, the problem is that I’m using a categorical classifier (the one on
>> the tutorial, LinearCSVM), I believe there should be a better one for
>> continuous data. Does that make any sense?
>>
>>
>> Thanks! And sorry for the newbie question
>>
>>
>>
>> _______________________________________________
>> Pkg-ExpPsy-PyMVPA mailing list
>> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
>> http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa
>>
>>
> --
> Joset A. Etzel, Ph.D.
> Research Analyst
> Cognitive Control & Psychopathology Lab
> Washington University in St. Louis
> http://mvpa.blogspot.com/
>
> _______________________________________________
> Pkg-ExpPsy-PyMVPA mailing list
> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/pkg-exppsy-pymvpa/attachments/20160218/5618477a/attachment.html>