Skip to content

Detail of publication


Šimandl, M. and Punčochář, I. : Using future information to change detection of stochastic systems . 5th international carpathian control conference, p. 659-664, AGH-UST, Krakow, 2004.


Change detection approaches of discrete stochastic systems are discussed in this paper. Many methods exploit a model of system and all available information up to the current time to make decision about parameters of system. Those methods do not consider influence of such decision on the future behavior of system to improve detection. The goal of this paper is to present an approach based on so called closed loop strategy used in optimal control. The approach respects influence of the future information on current decision. An appropriate loss function which fulfills designer requirements is minimized recursively using Bellman principle of optimality. Decrease in average of loss is obtained in comparison with the standard Bayesian approach.

Detail of publication

Title: Using future information to change detection of stochastic systems
Author: Šimandl, M. ; Punčochář, I.
Language: English
Date of publication: 25 May 2004
Year: 2004
Type of publication: Papers in proceedings of reviewed conferences
Title of journal or book: 5th international carpathian control conference
Page: 659 - 664
ISBN: 83-89772-00-0
Publisher: AGH-UST
Address: Krakow
Date: 25 May 2004 - 28 May 2004
/ /


change detection, stochastic system, Bellman principle, multiple models


 author = {\v{S}imandl, M. and Pun\v{c}och\'{a}\v{r}, I.},
 title = {Using future information to change detection of stochastic systems},
 year = {2004},
 publisher = {AGH-UST},
 journal = {5th international carpathian control conference},
 address = {Krakow},
 pages = {659-664},
 ISBN = {83-89772-00-0},
 url = {},