Subprocess-Constrained Markov Decision Processes
Constrained Markov decision processes (CMDPs) typically bound expected cumulative cost from an initial state. Such a guarantee can be insufficient when a policy must remain feasible after every history it may encounter. We introduce \emph{subprocess-constrained MDPs} (S-CMDPs), which require the expected discounted cost of each continuation, conditional on any reachable history, to satisfy a bound determined by the current state. This stronger requirement changes the structure and complexity of the problem: we show that stationary policies can be strictly suboptimal, and finding even an approximately optimal stationary policy is NP-hard. We adapt a recently introduced value-set iteration framework to show that history-dependent policies surprisingly overcome both obstacles. Our polynomial-time algorithm represents such a policy implicitly: its expected reward is within $ε$ of the optimal feasible value, it violates each subprocess cost bound by at most $δ$, and its action distribution can be computed in polynomial time for any given history.
Authors
- Jiarui Gan (ORCID: https://orcid.org/0000-0003-1280-2134)
- Debmalya Mandal (ORCID: https://orcid.org/0000-0002-8132-188X)
Publication Details
- Journal
- arXiv (Cornell University)
- Published
- 2026-11-16
- DOI
- https://doi.org/10.48550/arxiv.2611.21471
- Primary Topic
- Reinforcement Learning in Robotics
- Type
- preprint