Solving Long-run Average Reward Robust MDPs via Stochastic Games

no code implementations • 21 Dec 2023 • Krishnendu Chatterjee, Ehsan Kafshdar Goharshady, Mehrdad Karrabi, Petr Novotný, Đorđe Žikelić

First, we derive new computational complexity bounds for solving long-run average reward polytopic RMDPs, showing for the first time that the threshold decision problem for them is in NP coNP and that they admit a randomized algorithm with sub-exponential expected runtime.

Decision Making Decision Making Under Uncertainty