When do discounted-optimal policies also optimize the gain?
Systems and Control
2023-04-18 v1 Systems and Control
Optimization and Control
Abstract
In this technical note, we establish an upper-bound on the threshold on the discount factor starting from which all discounted-optimal deterministic policies are gain-optimal, that we prove to be tight on an example. To address computability issues of that theoretical threshold, we provide a weaker bound which is tractable on ergodic MDPs in polynomial time.
Cite
@article{arxiv.2304.08048,
title = {When do discounted-optimal policies also optimize the gain?},
author = {Victor Boone},
journal= {arXiv preprint arXiv:2304.08048},
year = {2023}
}