Cross-validation is fine even in this case, you'll just need to rearrange your
data in a way to leave-a-set-of-stimuli out, instead of leave-one-run-out.
Perhaps PyMVPA has some functionality to do this.
- Rawi
>________________________________
> From: Jacob Itzhacki <[email protected]>
>To: [email protected]
>Sent: Wednesday, October 24, 2012 1:11 PM
>Subject: [pymvpa] Alternatives to Cross-validation
>
>
>Dear all,
>
>
>First off thank you for your contribution. Great tool.
>
>
>Now that I am properly subscribed to the mailing list, I would like to ask:
>Is there is an alternative to cross-validation as I am hoping to use MVPA to
>train with one set of stimuli and test with a different one.
>I have been looking through the documentation but it is proving a difficult
>task as it is difficult to come up with an alternative keyword to search for.
>
>
>
>
>Thanks in advance,
>
>
>J. Itzhacki
>_______________________________________________
>Pkg-ExpPsy-PyMVPA mailing list
>[email protected]
>http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa
>
>
_______________________________________________
Pkg-ExpPsy-PyMVPA mailing list
[email protected]
http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-exppsy-pymvpa
