SAAC: Safe Reinforcement Learning as an Adversarial Game of Actor-Critics

Abstract

Although Reinforcement Learning (RL) is effective for sequential decision-making problems under uncertainty, it still fails to thrive in real-world systems where risk or safety is a binding constraint. In this paper, we formulate the RL problem with safety constraints as a non-zero-sum game. While deployed with maximum entropy RL, this formulation leads to a safe adversarially guided soft actor-critic framework, called SAAC. In SAAC, the adversary aims to break the safety constraint while the RL agent aims to maximize the constrained value function given the adversary's policy. The safety constraint on the agent's value function manifests only as a repulsion term between the agent's and the adversary's policies. Unlike previous approaches, SAAC can address different safety criteria such as safe exploration, mean-variance risk sensitivity, and CVaR-like coherent risk sensitivity. We illustrate the design of the adversary for these constraints. Then, in each of these variations, we show the agent differentiates itself from the adversary's unsafe actions in addition to learning to solve the task. Finally, for challenging continuous control tasks, we demonstrate that SAAC achieves faster convergence, better efficiency, and fewer failures to satisfy the safety constraints than risk-averse distributional RL and risk-neutral soft actor-critic algorithms. * This work was done during Yannis' PhD at Inria Lille (Scool team).

Malware behavior detection method based on reinforcement learning

“…Classification. Because of this analysis operation based on sequence features, it is consistent with reinforcement learning applied to sequence decision features 8 . This kind of problem that requires step-by-step operations on sequence data can be solved using policy gradients in reinforcement learning.…”

Multi-task safe reinforcement learning for navigating intersections in dense traffic

“…Safe RL. Constrained optimization techniques are usually adopted to solve safe RL problems (Garcıa & Fernández, 2015;Sootla et al, 2022;Yang et al, 2021;Flet-Berliac & Basu, 2022). Lagrangian-based methods use a multiplier to penalize constraint violations (Chow et al, 2017;Tessler et al, 2018;Stooke et al, 2020;Chen et al, 2021b).…”