A linear programming approach for the analysis of stochastic dynamic models (2000-2005)

The project is based on joint research of Helmes & Stockbridge, cf. NSF-Grants, DMS-9404990 and DMS-9803490. The objective is to develop efficient numerical tools for analyzing, among other problems, controlled and uncontrolled Markov processes. Moreover, the methods will be used to solve numerous dynamic as well as static optimization problems, e.g. zero-sum games on the unit square, exit-time and stopping problems for diffusion processes, special problems arising in mathematical finance global optimization problems, etc.