Improving the sample-efficiency of neural architecture search with reinforcement learning

Attila Nagy, Ábel Boros

Designing complex architectures has been an essential cogwheel in the revolution deep learning has brought about in the past decade. When solving difficult problems in a datadriven…