It is proved that there exist stationary optimal plans for discounted dynamic programming problems, and that there exist semi-Markov ϵ-optimal plans for positive dynamic programming problems. The actions are required to be taken in a variable action set F(s), and the reward function r(s, a) is a Borel measurable function of (s, a) and an u.s.c. function of a. Our results are related to recent work by Furukawa, Maitra, and Schal. The key tool is a generalization of a selection theorem of Dubins and Savage.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.