Tweeted By @rasbt

Also just added a function for computing the performance of any supervised model (classifier or regressor; assuming it has a fit+predict function) via the (pessimistic) out-of-bag bootstrap, the (optimistic) .632 Bootrap, and the .632+ Bootstrap: https://t.co/s62yswQkJV

— Sebastian Raschka (@rasbt) November 5, 2018