More `sklearn`-like API

[bnaul]

Currently we have a functional API with static methods like build_model.build_model_from_featureset and predict.model_predictions. Not only are these a bit of a mouthful, they're also a bit unintuitive because of how differently they are used than the corresponding sklearn functionality. I suggest that we try to modify our API to be more like that of sklearn, specifically:

1) Add a Model class that wraps a sklearn model and provides an sklearn-like API (.fit, .predict, .predict_proba) for interacting with Featureset objects. 2) (optional) Also change featurize to behave more like an sklearn preprocessor; maybe you'd initialize a Featureset and call .transform on a dataset? This is less obvious but there's probably some improvement to be made here as well.

[stefanv]

I am very much in favor of this proposal.

[acrellin]

Me too, and would like to hear more about what you have in mind regarding point 2. :+1:

[bnaul]

e.g. http://scikit-learn.org/stable/modules/feature_extraction.html

[arita37]

Additionnally, keeping api matching with Pandas and numpy would be a plus for better usage.