Dynamic system evolution and Markov chain approximation (Q1303524)

Mathematical modelling of dynamic system evolution is studied here as a problem in information theory. The construction of mathematical models is treated as a decision making process with limited information, and computational models for evolution are derived using the Markov chain approximations. Particularly, a stable approximation for the hyperbolic model is obtained and an algorithm is given. Some closely related results of the author on the associated stochastic optimal control are published somewhere else [cf. IEEE Int. Symp. Information Theory, Ulm; Germany, June 29--July 4 (1997), IEEE Catalog No. 97CH36074; Numer. Funct. Anal. Optimization 18, No. 3-4, 401-426 (1997; Zbl 0879.49006)].