New optimality conditions for average-payoff continuous-time Markov games in Polish spaces (Q547403)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | New optimality conditions for average-payoff continuous-time Markov games in Polish spaces |
scientific article |
Statements
New optimality conditions for average-payoff continuous-time Markov games in Polish spaces (English)
0 references
1 July 2011
0 references
The paper concerns two-person zero-sum continuous-time Markov games in which the state process evolves according to a continuous-time jump Markov process. The players can select their actions continuously in time, the state space and the action spaces for both players are all Polish spaces (that is, complete and separable metric spaces) and the expected average payoff criterion is studied. The transition rates are allowed to be unbounded, and the payoff function may have neither upper nor lower bounds. As is well-known, to prove the existence of a pair of optimal strategies for a zero-sum game problem, the optimality equation plays a key role. In this paper, two optimality inequalities are used to replace the optimality equation. Under more general conditions, these optimality inequalities yield the existence of the value of the game and of a pair of optimal stationary strategies. Under some additional conditions the optimality equation itself is further established. The paper contains several examples to illustrate the results, and also to show the difference between the conditions in this paper and those in the literature. In particular, one of these examples shows that considered approach allows nonergodic Markov processes.
0 references
zero-sum stochastic Markov games
0 references
average payoffs
0 references
optimality inequalities
0 references
unbounded transition rates
0 references
0 references
0 references
0 references
0 references
0 references
0 references
0 references