Efficient Bayesian Methods for Counting Processes in Partially Observable Environments

Ferdian Jovan, Jeremy Wyatt, Nick Hawes
Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, PMLR 84:1906-1913, 2018.

Abstract

When sensors that count events are unreliable, the data sets that result cannot be trusted. We address this common problem by developing practical Bayesian estimators for a partially observable Poisson process (POPP). Unlike Bayesian estimation for a fully observable Poisson process (FOPP) this is non-trivial, since there is no conjugate density for a POPP and the posterior has a number of elements that grow exponentially in the number of observed intervals. We present two tractable approximations, which we combine in a switching filter. This switching filter enables efficient and accurate estimation of the posterior. We perform a detailed empirical analysis, using both simulated and real-world data.

Cite this Paper

BibTeX
@InProceedings{pmlr-v84-jovan18a, title = {Efficient Bayesian Methods for Counting Processes in Partially Observable Environments}, author = {Jovan, Ferdian and Wyatt, Jeremy and Hawes, Nick}, booktitle = {Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics}, pages = {1906--1913}, year = {2018}, editor = {Storkey, Amos and Perez-Cruz, Fernando}, volume = {84}, series = {Proceedings of Machine Learning Research}, month = {09--11 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v84/jovan18a/jovan18a.pdf}, url = {https://proceedings.mlr.press/v84/jovan18a.html}, abstract = {When sensors that count events are unreliable, the data sets that result cannot be trusted. We address this common problem by developing practical Bayesian estimators for a partially observable Poisson process (POPP). Unlike Bayesian estimation for a fully observable Poisson process (FOPP) this is non-trivial, since there is no conjugate density for a POPP and the posterior has a number of elements that grow exponentially in the number of observed intervals. We present two tractable approximations, which we combine in a switching filter. This switching filter enables efficient and accurate estimation of the posterior. We perform a detailed empirical analysis, using both simulated and real-world data.} }
Endnote
%0 Conference Paper %T Efficient Bayesian Methods for Counting Processes in Partially Observable Environments %A Ferdian Jovan %A Jeremy Wyatt %A Nick Hawes %B Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2018 %E Amos Storkey %E Fernando Perez-Cruz %F pmlr-v84-jovan18a %I PMLR %P 1906--1913 %U https://proceedings.mlr.press/v84/jovan18a.html %V 84 %X When sensors that count events are unreliable, the data sets that result cannot be trusted. We address this common problem by developing practical Bayesian estimators for a partially observable Poisson process (POPP). Unlike Bayesian estimation for a fully observable Poisson process (FOPP) this is non-trivial, since there is no conjugate density for a POPP and the posterior has a number of elements that grow exponentially in the number of observed intervals. We present two tractable approximations, which we combine in a switching filter. This switching filter enables efficient and accurate estimation of the posterior. We perform a detailed empirical analysis, using both simulated and real-world data.
APA
Jovan, F., Wyatt, J. & Hawes, N.. (2018). Efficient Bayesian Methods for Counting Processes in Partially Observable Environments. Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 84:1906-1913 Available from https://proceedings.mlr.press/v84/jovan18a.html.

Related Material