Safe Policy Search for Lifelong Reinforcement Learning with Sublinear Regret.

Haitham Bou-Ammar, Rasul Tutunov, Eric Eaton

Browse the full ICML paper archive.