Semi-Markov decision processes with a reachable state-subset (Q3833895)
From MaRDI portal
!
This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:
scientific article; zbMATH DE number 4110497
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Semi-Markov decision processes with a reachable state-subset |
scientific article; zbMATH DE number 4110497 |
Statements
Semi-Markov decision processes with a reachable state-subset (English)
0 references
1989
0 references
unbounded costs
0 references
long-run average expected cost
0 references
semi-Markov decision process
0 references
general state and action spaces
0 references
optimality equation
0 references
\(\epsilon \) -optimal stationary policies
0 references
0 references
0 references
0 references
0 references
0.8630704283714294
0 references
0.8441866636276245
0 references
0.840437114238739
0 references
0.8342859148979187
0 references
0.83054518699646
0 references
The paper addresses the problem of minimising the long-run average expected cost per unit time in a semi-Markov decision process (SMDP) with general state and action spaces. The existence of solutions to the optimality equation, and thence of \(\epsilon\)-optimal stationary policies, is established under conditions which avoid standard continuity, compactness and boundedness assumptions, but allow contraction methods to be applied to an embedded SMDP. The work generalizes that of \textit{S. Lippman} [Ann. Math. Stat. 42, 1717-1726 (1971; Zbl 0231.90057)].
0 references