Also, it's not that GridSearch is sensitive in itself, but remember you're doing LeaveOneOut, so for every grid point you are actually doing `n_samples` calls to clf.fit. Maybe one of these calls is significantly slower than others due to scaling.
On Wed, Jul 3, 2013 at 10:42 PM, Lars Buitinck <l.j.buiti...@uva.nl> wrote: > 2013/7/3 Josh Wasserstein <ribonucle...@gmail.com>: >> Hmm, I noticed that if I run >> >> from sklearn import preprocessing >> X = preprocessing.scale(X) >> >> beforehand, it runs extremely fast! >> >> Why is that? > > Because support vector machines are quite sensitive to extreme feature > values. You should always scale anything that goes into an SVM. This > is documented: > http://scikit-learn.org/stable/modules/svm.html#tips-on-practical-use > > -- > Lars Buitinck > Scientific programmer, ILPS > University of Amsterdam > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by Windows: > > Build for Windows Store. > > http://p.sf.net/sfu/windows-dev2dev > _______________________________________________ > Scikit-learn-general mailing list > Scikit-learn-general@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/scikit-learn-general ------------------------------------------------------------------------------ This SF.net email is sponsored by Windows: Build for Windows Store. http://p.sf.net/sfu/windows-dev2dev _______________________________________________ Scikit-learn-general mailing list Scikit-learn-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/scikit-learn-general
