Maximum Entropy RL (Provably) Solves Some Robust RL Problems
Benjamin Eysenbach, Sergey Levine · arXiv (Cornell University) · 2021
Many potential applications of reinforcement learning (RL) require guarantees that the agent will perform well in the face of disturbances to the dynamics or reward function. In this paper, we prove theoretically that maximum entropy (MaxEnt) RL maximizes a lower bound on a robust RL objective, and thus can be used to learn policies that are robust to some disturbances in the dynamics and the reward function. While this capability of MaxEnt RL has been observed empirically in prior work, to the best of our knowledge our work provides the first rigorous proof and theoretical characterization of the MaxEnt RL robust set. While a number of prior robust RL algorithms have been designed to handle similar disturbances to the reward function or dynamics, these methods typically require additional moving parts and hyperparameters on top of a base RL algorithm. In contrast, our results suggest that MaxEnt RL by itself is robust to certain disturbances, without requiring any additional modifications. While this does not imply that MaxEnt RL is the best available robust RL method, MaxEnt RL is a simple robust RL method with appealing formal guarantees.
Cited by 27 works. doi:10.48550/arxiv.2103.06257 Open-access PDF
Authors
References
- Human-level control through deep reinforcement learning
- Markov Decision Processes: Discrete Stochastic Dynamic Programming.
- Robust and Optimal Control
- Continuous control with deep reinforcement learning
- Feedback Control Theory
- Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
- Agnostic Learning with Unknown Utilities
- Learning Continuous Control Policies by Stochastic Value Gradients
- Online Learning and Online Convex Optimization
- Function Optimization using Connectionist Reinforcement Learning Algorithms