Tag: Pedro Domingos

One of the most important problems in machine learning—and life—is the exploration-exploitation dilemma. If you’ve found something that works, should you just keep doing it? Or is it better to …