On Tue, Jul 30, 2013 at 09:38:15PM -0400, Josh Wasserstein wrote:
> but does scikit currently support classification/regression with missing
> features?
The upcoming 0.14 release (right now only available as an alpha) has a
simple imputation method:
http://scikit-learn.org/dev/modules/preprocessi
Sorry if this has been asked before, but does scikit currently support
classification/regression with missing features? (e.g. working with
something like NaN entries missing in the feature matrix).
If not, are there any proxies to leverage part of the data? (e.g. say I
want to use SVMs and that my
