Timezone: »

Online Regret Bounds for Undiscounted Continuous Reinforcement Learning
Ronald Ortner · Daniil Ryabko

Thu Dec 06 02:00 PM -- 12:00 AM (PST) @ Harrah’s Special Events Center 2nd Floor

We derive sublinear regret bounds for undiscounted reinforcement learning in continuous state space. The proposed algorithm combines state aggregation with the use of upper confidence bounds for implementing optimism in the face of uncertainty. Beside the existence of an optimal policy which satisfies the Poisson equation, the only assumptions made are Hoelder continuity of rewards and transition probabilities.

Author Information

Ronald Ortner (Montanuniversitaet Leoben)
Daniil Ryabko (INRIA)

More from the Same Authors