Factored Filtering of Continuous-Time Systems

E. Busra Celikkaya, Christian R. Shelton, William Lam
Proceedings of the 27th Conference on Uncertainty in Artificial Intelligence, PMLR R9:89-96, 2011.

Abstract

We consider filtering for a continuous-time, or asynchronous, stochastic system where the full distribution over states is too large to be stored or calculated. We assume that the rate matrix of the system can be compactly represented and that the belief distribution is to be approximated as a product of marginals. The essential computation is the matrix exponential. We look at two different methods for its computation: ODE integration and uniformization of the Taylor expansion. For both we consider approximations in which only a factored belief state is maintained. For factored uniformization we demonstrate that the KL-divergence of the filtering is bounded. Our experimental results confirm our factored uniformization performs better than previously suggested uniformization methods and the mean field algorithm.

Cite this Paper

BibTeX
@InProceedings{pmlr-vR9-celikkaya11a, title = {Factored Filtering of Continuous-Time Systems}, author = {Celikkaya, E. Busra and Shelton, Christian R. and Lam, William}, booktitle = {Proceedings of the 27th Conference on Uncertainty in Artificial Intelligence}, pages = {89--96}, year = {2011}, editor = {Cozman, Fabio and Pfeffer, Avi}, volume = {R9}, series = {Proceedings of Machine Learning Research}, month = {14--17 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/r9/main/assets/celikkaya11a/celikkaya11a.pdf}, url = {https://proceedings.mlr.press/r9/celikkaya11a.html}, abstract = {We consider filtering for a continuous-time, or asynchronous, stochastic system where the full distribution over states is too large to be stored or calculated. We assume that the rate matrix of the system can be compactly represented and that the belief distribution is to be approximated as a product of marginals. The essential computation is the matrix exponential. We look at two different methods for its computation: ODE integration and uniformization of the Taylor expansion. For both we consider approximations in which only a factored belief state is maintained. For factored uniformization we demonstrate that the KL-divergence of the filtering is bounded. Our experimental results confirm our factored uniformization performs better than previously suggested uniformization methods and the mean field algorithm.}, note = {Reissued by PMLR on 04 October 2026.} }
Endnote
%0 Conference Paper %T Factored Filtering of Continuous-Time Systems %A E. Busra Celikkaya %A Christian R. Shelton %A William Lam %B Proceedings of the 27th Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2011 %E Fabio Cozman %E Avi Pfeffer %F pmlr-vR9-celikkaya11a %I PMLR %P 89--96 %U https://proceedings.mlr.press/r9/celikkaya11a.html %V R9 %X We consider filtering for a continuous-time, or asynchronous, stochastic system where the full distribution over states is too large to be stored or calculated. We assume that the rate matrix of the system can be compactly represented and that the belief distribution is to be approximated as a product of marginals. The essential computation is the matrix exponential. We look at two different methods for its computation: ODE integration and uniformization of the Taylor expansion. For both we consider approximations in which only a factored belief state is maintained. For factored uniformization we demonstrate that the KL-divergence of the filtering is bounded. Our experimental results confirm our factored uniformization performs better than previously suggested uniformization methods and the mean field algorithm. %Z Reissued by PMLR on 04 October 2026.
APA
Celikkaya, E.B., Shelton, C.R. & Lam, W.. (2011). Factored Filtering of Continuous-Time Systems. Proceedings of the 27th Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research R9:89-96 Available from https://proceedings.mlr.press/r9/celikkaya11a.html. Reissued by PMLR on 04 October 2026.

Related Material