Abstract:
Effective cases of filtering from jerky observations are considered. Conditions are given under which the a posteriori distribution of the characteristics of processes is Gaussian with a probability of one. Expressions for optimal estimates are given.