DeLTA seminar by Haipeng Luo
Speaker
Haipeng Luo, University of Southern California
Title
The best of both worlds: stochastic and adversarial episodic MDPs with unknown transition
Abstract
We consider the best-of-both-worlds problem for learning an episodic Markov Decision Process through $T$ episodes, with the goal of achieving $\widetilde{\mathcal{O}}(\sqrt{T})$ regret when the losses are adversarial and simultaneously $\mathcal{O}(\text{polylog}(T))$ regret when the losses are (almost) stochastic. Recent work by [Jin and Luo, 2020] achieves this goal when the fixed transition is known, and leaves the case of unknown transition as a major open question. In this work, we resolve this open problem by using the same Follow-the-Regularized-Leader (FTRL) framework together with a set of new techniques. Specifically, we first propose a loss-shifting trick in the FTRL analysis, which greatly simplifies the approach of [Jin and Luo, 2020] and already improves their results for the known transition case. Then, we extend this idea to the unknown transition case and develop a novel analysis which upper bounds the transition estimation error by (a fraction of) the regret itself in the stochastic setting, a key property to ensure $\mathcal{O}(\text{polylog}(T))$ regret.
_____________________________
Upcoming DeLTA seminars:
6 September 2021 @ 10:00. Ola Rønning.
13 September 2021 @ 10:00. Shinji Ito.
-------------------
You can subscribe to DeLTA Seminar mailing list by sending an empty email to delta-seminar-join@list.ku.dk.
Information about DeLTA Seminars is available at DeLTA Lab page: https://sites.google.com/diku.edu/delta
DeLTA is a research group affiliated with the Department of Computer Science at the University of Copenhagen studying diverse aspects of Machine Learning Theory and its applications, including, but not limited to Reinforcement Learning, Online Learning and Bandits, PAC-Bayesian analysis