One Risk to Rule Them All: A Risk-Sensitive Perspective on Model-Based Offline Reinforcement Learning

TL;DR

This paper introduces a risk-sensitive model-based offline reinforcement learning approach that accounts for both epistemic and aleatoric uncertainties, improving safety and performance in stochastic environments.

Contribution

It proposes a unified risk-sensitive framework that jointly addresses distributional shift and stochasticity in offline RL, using risk-averse modeling of uncertainties.

Findings

Achieves competitive results on deterministic benchmarks.

Outperforms existing risk-sensitive methods in stochastic domains.

Effectively prevents distributional shift and poor outcomes.

Abstract

Offline reinforcement learning (RL) is suitable for safety-critical domains where online exploration is too costly or dangerous. In such safety-critical settings, decision-making should take into consideration the risk of catastrophic outcomes. In other words, decision-making should be risk-sensitive. Previous works on risk in offline RL combine together offline RL techniques, to avoid distributional shift, with risk-sensitive RL algorithms, to achieve risk-sensitivity. In this work, we propose risk-sensitivity as a mechanism to jointly address both of these issues. Our model-based approach is risk-averse to both epistemic and aleatoric uncertainty. Risk-aversion to epistemic uncertainty prevents distributional shift, as areas not covered by the dataset have high epistemic uncertainty. Risk-aversion to aleatoric uncertainty discourages actions that may result in poor outcomes due to environment stochasticity. Our experiments show that our algorithm achieves competitive performance on deterministic benchmarks, and outperforms existing approaches for risk-sensitive objectives in stochastic domains.