IBM interview question
How to prevent over-fitting?
Interview Answers
Anonymous
12 Sept 2017
1) get more training data 2) using weight decay to limit capacity of the model 3) using early-stopping 4) using cross-validation 5) average many different models 6) drop-out
1