IPACS Electronic library

Distributions of Multiestimates for Statistically Uncertain Systems

Boris Ananiev
Problems of evolution describing are considered
for probability distributions (both unconditional and conditional) of random information sets. These sets named in the work as multiestimates naturally appear in problems of state and parameter estimation for statistically uncertain systems. The special attention is given to cases when multiestimates depend on a finite number of parameters. With use of theories of guaranteed and
statistical estimation for such cases, the estimates from below are received for probabilities of inclusion of a multiestimate in the given set or its intersection with this set. Results are illustrated on examples.
File: download
Copyright © 2003—2015 The Laboratory "Control of Complex Systems", IPME RAS