Approximations in Dynamic Zero-sum Games, Ii Approximations in Dynamic Zero-sum Games, Ii
نویسندگان
چکیده
We pursue in this paper our study of approximations of values and-saddle-point policies in dynamic zero-sum games. After extending the general theorem for approximation, we study zero-sum stochastic games with countable state space, and non-bounded immediate reward. We focus on the expected average payoo criterion. We use some tools developed in the rst paper, to obtain the convergence of the values as well as the convergence of the saddle-point policies in various approximation problems. We consider several schemes of truncation of the state space (e.g. nite state approximation) and approximations of games with discount factor close to one by the game with expected average cost. We use the extension of the general Theorem for approximation to study approximations in stochastic games with complete information. We nally consider the problem of approximating the sets of policies. We obtain some general results that we apply to a pursuit evasion diierential game. Approximations dans les jeux dynamiques a somme nulle, II R esum e : Nous poursuivons dans ce papier une etude portant sur l'approximation de la fonctions valeur, ainsi que des strategies-optimal pour des jeux dynamiques a deux joueurs et a somme nulle. Nous etendons dans un premier temps un th eor eme g en eral utilis e pour les approximations , puis nous etudions des jeux stochastiques a somme nulle dont l'espace d' etat est d enombrable, et dont le co^ ut instantan e est non born e. On s'interesse plus particuli erement au co^ ut moyen. Nous utilisons des outils que nous avons developp es dans le pr ec edent papier, pour obtenir la convergence de la fonction valeur ainsi que la convergence des strat egies-optimales dans dii erents probl emes d'approximation. Nous consid erons dii erents types d'approximations d'espaces d' etats innnis par des espaces d' etat nis ainsi que des approximations de jeux avec des taux d'actua-lisation proche de 1, par des jeux a co^ uts moyens. Nous utilisons l'extension du th eor eme g en eral pour les approximations pour etudier des approximations pour des jeux stochastiques avec information compl ete. On consid ere ennn le probl eme de l'approximation ni des ensembles de stra-t egies. Nous obtenons des r esultats g en eraux que nous appliquons au cas de jeux dii erentiels de poursuite-evasion.
منابع مشابه
Approximations in Dynamic Zero-sum Games, I
We develop a unifying approach for approximating a \limit" zero-sum game by a sequence of approximating games. We discuss both the convergence of the values and the convergence of optimal (or \almost" optimal) strategies. Moreover, based on optimal policies for the limit game, we construct policies which are almost optimal for the approximating games. We then apply the general framework to stat...
متن کاملA TRANSITION FROM TWO-PERSON ZERO-SUM GAMES TO COOPERATIVE GAMES WITH FUZZY PAYOFFS
In this paper, we deal with games with fuzzy payoffs. We proved that players who are playing a zero-sum game with fuzzy payoffs against Nature are able to increase their joint payoff, and hence their individual payoffs by cooperating. It is shown that, a cooperative game with the fuzzy characteristic function can be constructed via the optimal game values of the zero-sum games with fuzzy payoff...
متن کاملApproximations in Dynamic Zero-sum Games
We develop a unifying approach for approximating a “limit" zero-sum game by a sequence of approximating games. We discuss both the convergence of the values and the convergence of optimal (or “almost" optimal) strategies. Moreover, based on optimal policies for the limit game, we construct policies which are almost optimal for the approximating games. We then apply the general framework to stat...
متن کاملStochastic Differential Games and Intricacy of Information Structures
This paper discusses, in both continuous time and discrete time, the issue of certainty equivalence in two-player zero-sum stochastic differential/dynamic games when the players have access to state information through a common noisy measurement channel. For the discrete-time case, the channel is also allowed to fail sporadically according to an independent Bernoulli process, leading to intermi...
متن کاملZero-Sum Repeated Games: Recent Advances and New Links with Differential Games
The purpose of this survey is to describe some recent advances in zero-sum repeated games and in particular new connections to differential games. Topics include: approachability, asymptotic analysis: recursive formula and operator approach, dual game and incomplete information, uniform approach.
متن کامل