Efficient Risk-Averse Reinforcement Learning
This addresses the issue of local optima in risk-averse RL for applications such as autonomous systems, though it is incremental as it builds on existing policy gradient methods.
The paper tackled the problem of risk-averse reinforcement learning methods ignoring high-return strategies, which leads to local optima, by proposing a soft risk mechanism and a Cross Entropy module for risk sampling, resulting in improved risk aversion in benchmarks like maze navigation and autonomous driving where standard methods fail.
In risk-averse reinforcement learning (RL), the goal is to optimize some risk measure of the returns. A risk measure often focuses on the worst returns out of the agent's experience. As a result, standard methods for risk-averse RL often ignore high-return strategies. We prove that under certain conditions this inevitably leads to a local-optimum barrier, and propose a soft risk mechanism to bypass it. We also devise a novel Cross Entropy module for risk sampling, which (1) preserves risk aversion despite the soft risk; (2) independently improves sample efficiency. By separating the risk aversion of the sampler and the optimizer, we can sample episodes with poor conditions, yet optimize with respect to successful strategies. We combine these two concepts in CeSoR - Cross-entropy Soft-Risk optimization algorithm - which can be applied on top of any risk-averse policy gradient (PG) method. We demonstrate improved risk aversion in maze navigation, autonomous driving, and resource allocation benchmarks, including in scenarios where standard risk-averse PG completely fails.