Zero-Sum Stochastic Games with Partial Information and Average Payoff

TL;DR

This paper analyzes partially observable zero-sum stochastic games with average payoff, demonstrating the existence of a game value and deriving optimal strategies through an equivalent fully observable game approach.

Contribution

It introduces a method to analyze partially observable zero-sum stochastic games with average payoff by transforming them into fully observable games, establishing the existence of a value and optimal strategies.

Findings

The game has a well-defined value.

Optimal strategies for both players are characterized.

The approach uses an equivalent fully observable game model.

Abstract

We consider discrete time partially observable zero-sum stochastic game with average payoff criterion. We study the game using an equivalent completely observable game. We show that the game has a value and also we come up with a pair of optimal strategies for both the players.