Re: [Scikit-learn-general] Classification if only one class is present

I agree with you, Andy. It seems the right thing to do to address Olivier & Gilles' subsampling concern is to fix the subsampling, not allow the ensemble to add useless classifiers/do useless cross-validation folds. On Fri, Aug 10, 2012 at 7:48 AM, Andreas Mueller wrote: > No-one has any opinion

Re: [Scikit-learn-general] Classification if only one class is present

No-one has any opinion on that? On 08/05/2012 10:01 PM, Andreas Mueller wrote: Hey everybody. Recently I opened and issue suggesting that all classifiers should raise an error if there is only one class present during training. My rea

Re: [Scikit-learn-general] No methods seem to predict well

Zach, can you check the construction of your test set - The CV scores on the training set are OK but when I apply the model to the test set it fails miserably - looking at the feature distribution between training and test set indicates that there's something fishy but its hard to tell given the s