[pymvpa] cross-validation & regression analysis

Matthias Ekman Matthias.Ekman at nf.mpg.de
Thu Apr 8 23:33:26 UTC 2010


Yaroslav Halchenko wrote:
> look into doc for TransferError and mvpa/tests/test_regr.py for
> "examples" ;-)
> 
> PS I thought we had switching between MeanMismatch for classification
> and CorrError for regression... my bad -- need to do that for
> TransferError

nice -- thanks for the helpfull pointer!

For some reasons CorrErrorFx is not listed in the "old" online 
documentation:
http://www.pymvpa.org/modref/mvpa.misc.errorfx.html?highlight=errorfx#module-mvpa.misc.errorfx

Matthias


> 
> On Fri, 09 Apr 2010, Matthias Ekman wrote:
> 
>> Hi,
> 
>> I just started to figure out how SVM for 'regression' is implemented
>> in PyMVPA/libsvm. In comparison to SVM for 'classification' I am not
>> interested in the classification accuracy but in the explained
>> variance.
> 



More information about the Pkg-ExpPsy-PyMVPA mailing list