PROJECT TITLE:

An Optimizer's Approach to Stochastic Control Problems With Nonclassical Information Structures

ABSTRACT:

We have a tendency to present a general optimization-based framework for stochastic management issues with nonclassical info structures. We have a tendency to cast these issues equivalently as optimization problems on joint distributions. The resulting problems are essentially nonconvex. Our approach to solving them is thru convex relaxation . We solve the instance solved by Bansal and Başar (“Stochastic teams with nonclassical information revisited: When is an affine law optimal?”, IEEE Trans. Automatic Management, 1987) with a specific application of this approach that uses the information processing inequality for constructing the convex relaxation. Using bound $f$-divergences, we tend to obtain a new, larger set of inverse optimal price functions for such problems. Insights are obtained on the relation between the structure of cost functions and of convex relaxations for inverse optimal control.