Online Non-Convex Learning: Following the Perturbed Leader is Optimal.

Arun Sai Suggala, Praneeth Netrapalli

Browse the full ALT paper archive.