[pymvpa] Use searchlight algorithm only for predictions.

[J.A. Etzel]

One strategy is to use multiple classification analyses and compare the 
results; sort of like the 'virtual lesion' method. Using your flag 
example, a classifier given only "has stripes" - the "has stars" feature 
was lesioned - will fail, but one given both 'has stars' and 'has 
stripes' will succeed. In the context of fMRI, classification could 
succeed when an entire ROI is used, but fail when a subpart of the ROI 
is omitted.

This type of strategy is not always applicable (or straightforward to 
implement and interpret), but sounds to me like it might be a bit closer 
to what you're trying to find out.

Jo



On 1/31/2013 8:53 AM, Roberto Guidotti wrote:
> Thank you guys,
>
> Yes, I know that I can't predict using a portion of voxel. Let's say
> that I would like to train on full brain and test on a portion, putting
> out of ROI voxel intensity to zero.
> I don't know if it makes sense conceptually because I would like to
> predict using a portion of features on a model built on multiple features.
>
> Probably could be an sensitivity measure, e.g. building a classifier to
> predict from flag the country for example Liberarian flag and stars and
> stripes, If I use features from stripes part of the flag (common in both
> of the flags) the classifier isn't able to classify - well, using a
> feature selection probably those features were discarded - but using
> "stars" part as ROI the classifier identifies the flag, and so I will
> know where the classifier is more sensitive! (Hope I explained it clear).
>
> I don't know if there is still a theoretical problem.
>
> Thank you
> R
>