References
Facsimile [PDF, 4.9M]

Saddle Point Conditions for Antagonistic Positional Games in Complex Markov Decision Processes

Dmitrii Lozovanu and Stefan Pickl

p. 105-116

    Abstract

    A class of stochastic antagonistic positional games for Markov decision processes with average and expected total discounted costs optimization criteria are formu lated and studied. Saddle point conditions in the considered class of games that extend saddle point conditions for deterministic parity games are derived. Further more algorithms for determining the optimal stationary strategies of the players are proposed and grounded

    Index

    Keywords

    Markov decision processes, stochastic positional games, antagonistic positional games, saddle point, optimal stationary strategies

    Text

    Download Facsimile [PDF, 4.9M]

    References

    Bibliographical reference

    Dmitrii Lozovanu and Stefan Pickl, « Saddle Point Conditions for Antagonistic Positional Games in Complex Markov Decision Processes », CASYS, 30 | 2014, 105-116.

    Electronic reference

    Dmitrii Lozovanu and Stefan Pickl, « Saddle Point Conditions for Antagonistic Positional Games in Complex Markov Decision Processes », CASYS [Online], 30 | 2014, Online since 17 October 2024, connection on 14 November 2024. URL : http://popups.lib.uliege.be/1373-5411/index.php?id=4641

    Authors

    Dmitrii Lozovanu

    Institute of Mathematics and Computer Science, Academy of Sciences of Moldova, Academy str., 5, Chisinau, MD-2028, Moldova

    By this author

    Stefan Pickl

    Institute for Theoretical Computer Science, Mathematics and Operations Research, Universität der Bundeswehr München 85577 Neubiberg-München, Germany

    By this author

    Copyright

    CC BY-SA 4.0 Deed