[pymvpa] Example Dataset

[Per B. Sederberg]

Well, it looks like the data labels ARE shifted.  When running a
Bottle vs. Scissors comparison I got 50% accuracy when shifted by 3
and 58% correct when not shifted.

Not convinced by that I ran a Face vs. House classification and got
99.5% accuracy with no shift, but only 76.9% accuracy when shifted
by 3.

All these tests were with a Linear SVM classifier, so they should be
stable results.

Even though all signs are pointing to the fact that they were shifted,
perhaps Michael knows for sure whether they were...

Have a great night,
Per


At Tue, 13 Jan 2009 22:21:30 -0500,
Per B. Sederberg wrote:
> 
> Okeydokey!  I'll try it out and let you know if the generalization
> performance improves.  It's already pretty darn good for most classes,
> but I'll focus on a hard one.
> 
> Thanks,
> P
> 
> At Tue, 13 Jan 2009 22:00:25 -0500,
> Yaroslav Halchenko wrote:
> > 
> > I guess it wouldn't be too hard to figure it out empirically (atm I
> > can't recall what was actually done -- I think indeed they are not
> > corrected for HRF since Haxby was interested in block averages):
> > just shift labels by 1,2,3 samples and see when the generalization
> > performance is the best ;) should be 1 line addition to smth like
> > http://www.pymvpa.org/ex/start_easy.html
> > 
> > On Tue, 13 Jan 2009, Per B. Sederberg wrote:
> > 
> > > Hi Everybody:
> > 
> > > I just wanted to check to see whether the labeled samples (listed in
> > > attributes_literal.txt) in the Haxby et al. (2001) example subject are
> > > shifted to the peak hemodynamic response or if one must perform some
> > > type of shifting (i.e., convolving with HRF) to line up the samples
> > > with the presented image.
> > 
> > > Although I assume they are shifted because none of the examples
> > > perform this step to line up the labels with one would expect the peak
> > > brain activity, I was unable to confirm this from the download page:
> > > http://www.pymvpa.org/examples.html#exampledata
> > 
> > > Thanks,
> > > Per
> > 
> > > _______________________________________________
> > > Pkg-ExpPsy-PyMVPA mailing list
> > > Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> > > http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa
> > 
> > 
> > -- 
> > 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        
> > 
> > _______________________________________________
> > Pkg-ExpPsy-PyMVPA mailing list
> > Pkg-ExpPsy-PyMVPA at lists.alioth.debian.org
> > http://lists.alioth.debian.org/mailman/listinfo/pkg-exppsy-pymvpa