Timezone: »

 
Poster
Unifying PAC and Regret: Uniform PAC Bounds for Episodic Reinforcement Learning
Christoph Dann · Tor Lattimore · Emma Brunskill

Wed Dec 06 06:30 PM -- 10:30 PM (PST) @ Pacific Ballroom #198
in Posters Wed »

Statistical performance bounds for reinforcement learning (RL) algorithms can be critical for high-stakes applications like healthcare. This paper introduces a new framework for theoretically measuring the performance of such algorithms called Uniform-PAC, which is a strengthening of the classical Probably Approximately Correct (PAC) framework. In contrast to the PAC framework, the uniform version may be used to derive high probability regret guarantees and so forms a bridge between the two setups that has been missing in the literature. We demonstrate the benefits of the new framework for finite-state episodic MDPs with a new algorithm that is Uniform-PAC and simultaneously achieves optimal regret and PAC guarantees except for a factor of the horizon.

Keywords: Reinforcement Learning and Planning · Exploration

Author Information

Christoph Dann (Carnegie Mellon University)
Tor Lattimore (DeepMind)
Emma Brunskill (Stanford University)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors