SMOTE sampling method for imbalanced data #11279
Description
Add [SMOTE|http://www.jair.org/papers/paper953.html], "Synthetic Minority Over-sampling Technique" for handling imbalanced datasets/ This is a more sophisticated means of balancing the dataset vs straight-forward over/undersampling that we currently have implemented in H2O via the balance_classes, class_sampling_factors, max_after_balance_size arguments.
Overview (see #4): http://machinelearningmastery.com/tactics-to-combat-imbalanced-classes-in-your-machine-learning-dataset/
We could also consider the [ROSE|https://www.openstarts.units.it/dspace/bitstream/10077/4002/1/Menardi%20Torelli%20DEAMS%20WPS2.pdf] method (available in the caret R package).