Provably Efficient Multi-Task Reinforcement Learning with Model Transfer

2021/07/19 by Zhang, Chicheng, Wang, Zhi · 1 citation
#FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.2107.08622

Abstract

We study multi-task reinforcement learning (RL) in tabular episodic Markov decision processes (MDPs). We formulate a heterogeneous multi-player RL problem, in which a group of players concurrently face similar but not necessarily identical MDPs, with a goal of improving their collective performance through inter-player information sharing. We design and analyze an algorithm based on the idea of model transfer, and provide gap-dependent and gap-independent upper and lower bounds that characterize the intrinsic complexity of the problem.

Cited by

• Reinforcement Learning for Individual Optimal Policy from Heterogeneous Data

Related

• Provable Benefits of Multi-task RL under Non-Markovian Decision Making Processes
• Sample Complexity of Multi-task Reinforcement Learning
• Reinforcement Learning in the Wild with Maximum Likelihood-based Model Transfer
• Transfer from Multiple MDPs
• Learning Shared Representations in Multi-task Reinforcement Learning