Risk-averse Total-reward MDPs with ERM and EVaR
This work addresses the problem of risk-averse reinforcement learning for researchers and practitioners, offering a more tractable approach compared to prior methods, though it is incremental in improving upon existing risk-averse MDP frameworks.
The paper tackles the challenge of optimizing risk-averse objectives in Markov Decision Processes (MDPs) by showing that the total reward criterion under ERM and EVaR risk measures can be optimized using stationary policies, enabling simpler analysis and deployment with methods like exponential value iteration.
Optimizing risk-averse objectives in discounted MDPs is challenging because most models do not admit direct dynamic programming equations and require complex history-dependent policies. In this paper, we show that the risk-averse {\em total reward criterion}, under the Entropic Risk Measure (ERM) and Entropic Value at Risk (EVaR) risk measures, can be optimized by a stationary policy, making it simple to analyze, interpret, and deploy. We propose exponential value iteration, policy iteration, and linear programming to compute optimal policies. Compared with prior work, our results only require the relatively mild condition of transient MDPs and allow for {\em both} positive and negative rewards. Our results indicate that the total reward criterion may be preferable to the discounted criterion in a broad range of risk-averse reinforcement learning domains.