arxiv7.0 / 10

Subprocess-Constrained Markov Decision Processes

Jiarui Gan, Debmalya Mandal

Abstract

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.

Research area

ai reliabilityguaranteed safe aireinforcement learning
Published
16 Nov 2026
Source
arxiv
Org
University of Oxford
View paper
Sign in to read and join the discussion.