[pymvpa] CRM-SPAM: [mvpa-toolbox] Quantification of classification decision

[Yaroslav Halchenko]

sorry for the shameless ad... in pymvpa [1] you can use

clf.values

to get 'raw' values prior to their transformation quantized decision
value (e.g. -1/+1). In SVM for binary problem it is simply the value you
desire...

[1] http://www.pymvpa.org/

On Fri, 07 Nov 2008, Jong Yoon wrote:

> Is it possible to extract a numeric value for the error value
> determining classifcation decision for each test scan?  Say the
> classifier learned to distinguish faces and scenes and then tested on
> separate scans for the classification of face patterns of activity,
> instead of just getting a record of the classifier's guesses, e.g.
> face or scene, is it possible to get a numeric value which determined
> whether the classifier guesses a face or a scene on a scan-byscan
> basis?   If so, where in the code can I retrieve this quantity?  I can
> think of a few applications in which such quantification of the
> goodness of fit so to speak can come in very handy.

> Thanks in advance.


> --~--~---------~--~----~------------~-------~--~----~
> You received this message because you are subscribed to the Google Groups "Princeton MVPA Toolbox for Matlab" group.
> To post to this group, send email to mvpa-toolbox at googlegroups.com
> To unsubscribe from this group, send email to mvpa-toolbox+unsubscribe at googlegroups.com
> For more options, visit this group at http://groups.google.com/group/mvpa-toolbox?hl=en
> -~----------~----~----~----~------~----~------~--~---



-- 
Yaroslav Halchenko
Research Assistant, Psychology Department, Rutgers-Newark
Student  Ph.D. @ CS Dept. NJIT
Office: (973) 353-1412 | FWD: 82823 | Fax: (973) 353-1171
        101 Warren Str, Smith Hall, Rm 4-105, Newark NJ 07102
WWW:     http://www.linkedin.com/in/yarik