[pymvpa] Questions about mvpa preprocessing...

[Roberto Guidotti]

Thank you all for your precious response,

I've done other test, following Micheal advice, and results seems to be very
strange, for my eyes! I've tried to classify Rest vs ClassOne and Rest vs
ClassTwo, well with GNB I've a good classification accuracy for Rest vs
ClassOne, or better to say a good confusion matrix, And with the other class
bad results... with SVM I have the opposite!
Seems to be that I have one state (ClassOne) that attracts other with SVM
and another state (Rest) that attract other with GNB, I've also a three
label classification.
Is dataset unclassifiable?

Thank you
RG

2011/4/16 Francisco Pereira <francisco.pereira at gmail.com>

> The only thing I can add to what Michael said is that you could look
> at the predictions for every example and see what resting state gets
> misclassified as or, conversely, what other things get misclassified
> as resting state.
>
> cheers,
> Francisco
>
> On Fri, Apr 15, 2011 at 1:45 PM, Roberto Guidotti <robbenson18 at gmail.com>
> wrote:
> > Thank you, for the quick response and for the advice.
> >
> > For the first question: in the review papers I've read this question is
> > never treated in depth, but only if the study is across-subject, thank
> you
> > for the paper you link to me. My study is preliminary, I want to check
> that
> > MVPA technique tells me something else... My analysis is firstly
> > within-subject and then I could "upgrade" to cross-subject!!
> >
> > The second question: I know it were incomprehensible ;) now I'll explain
> > simply :D I have a task with two conditions plus resting, without resting
> > state (fds = fds[fds.sa.targets != 'Rest']) my accuracy is good, with the
> > full dataset I have bad results (no. of training examples for each class
> are
> > balanced) I don't know if I the independence conditions, with this
> dataset,
> > could be true. My question is the fact that my resting state is not
> > classified as resting state, or my conditions are misclassified, is it an
> > alarm bell?
> >
> > Thank you.
> > Roberto
> >
> > 2011/4/15 J.A. Etzel <jetzel at artsci.wustl.edu>
> >>
> >> I've often performed MVPA after spatial normalization. As with most
> >> things, whether it's a good idea depends on your analysis goals: if you
> will
> >> perform an across-subjects analysis some sort of spatial normalization
> is
> >> required; if only within-subjects, it's optional.
> >>
> >> As a plug, I wrote about some of these issues in my tutorial paper:
> Etzel
> >> JA, Gazzola V, Keysers C. 2009. An introduction to anatomical ROI-based
> fMRI
> >> classification analysis. Brain Research 1282:114-125. There are quite a
> few
> >> other introductory papers, too - check the citations.
> >>
> >> I'd need more details for the resting state question. It can be very
> nice
> >> to include a "control" (should be chance) classification in an analysis.
> >>
> >> good luck,
> >> Jo
> >>
> >>
> >> On 4/15/2011 9:40 AM, Roberto Guidotti wrote:
> >>>
> >>> Dear all,
> >>>
> >>> I'm not sure if this is the right place to ask these questions:
> >>> Is it right to coregister fMRI images to standard atlas and then
> perform
> >>> the MVPA analysis? If yes, there are some other issue to take care in
> >>> the post analysis operation?
> >>> Is important that the resting state, if included in classification, has
> >>> bad classifier accuracy?
> >>>
> >>> I'm sorry if these are trivial question and if I'm off topic.
> >>>
> >>> Thank you
> >>> RG
> >>>
> >>>
> >>>
> >>> _______________________________________________
> >>> Pkg-ExpPsy-PyMVPA mailing list
> >>> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> >>> http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa
> >>
> >> _______________________________________________
> >> Pkg-ExpPsy-PyMVPA mailing list
> >> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> >> http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa
> >
> >
> > _______________________________________________
> > Pkg-ExpPsy-PyMVPA mailing list
> > Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> > http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa
> >
> >
>
> _______________________________________________
> Pkg-ExpPsy-PyMVPA mailing list
> Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.alioth.debian.org/pipermail/pkg-exppsy-pymvpa/attachments/20110420/8686509b/attachment.htm>