Accession Number:

ADA222409

Title:

Robust Approximations for the Filtering Problem

Descriptive Note:

Corporate Author:

ALBERTA UNIV EDMONTON DEPT OF STATISTICS AND APPLIED PROBABILITY

Personal Author(s):

Report Date:

1982-01-01

Pagination or Media Count:

5.0

Abstract:

A diffusion observation process is approximated by a Markov chain. The information obtained by observing the Markov chain is the same as that obtained by observing a related multivariate point process. Filtering and Zakai equations are obtained for multivariate point process observations. These involve Stieltjes integrals rather than Ito integrals with respect to Brownian motion, and so they provide robust formulae, that is, formulae which are continuous in the observation process. kr

Subject Categories:

  • Statistics and Probability

Distribution Statement:

APPROVED FOR PUBLIC RELEASE