State of system approximation for stochastic systems

This paper describes a new contribution to stochastic, nonlinear system approximation. An approach to approximating stochastic systems is described by which the difficulties associated with the erratic behavior of sample functions are circumvented. This new approximation criterion is called uniform in-probability approximation, where the probability of the absolute approximation error exceeding a prescribed /spl epsiv/>0 is uniformly less than a prescribed probability p.