Decision Problems with Multiple Probabilistic Models

This paper discusses stochastic multi-agent team problems wherein the decision makers have different probabilistic models of the underlying decision process. A suitable equilibrium solution concept is introduced for such decision problems which exhibit probabilistic multi-modeling, and the existence, uniqueness and stability properties of this equilibrium solution are studied under static information patterns. The special case of Gaussian distributions is studied in some depth, and some explicit equilibrium policies are derived for both discrete and continuous-time team problems.