Robust and Scalable Bayesian Online Changepoint Detection

Matias Altamirano, Francois-Xavier Briol, Jeremias Knoblauch
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:642-663, 2023.

Abstract

This paper proposes an online, provably robust, and scalable Bayesian approach for changepoint detection. The resulting algorithm has key advantages over previous work: it provides provable robustness by leveraging the generalised Bayesian perspective, and also addresses the scalability issues of previous attempts. Specifically, the proposed generalised Bayesian formalism leads to conjugate posteriors whose parameters are available in closed form by leveraging diffusion score matching. The resulting algorithm is exact, can be updated through simple algebra, and is more than 10 times faster than its closest competitor.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-altamirano23a, title = {Robust and Scalable {B}ayesian Online Changepoint Detection}, author = {Altamirano, Matias and Briol, Francois-Xavier and Knoblauch, Jeremias}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {642--663}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/altamirano23a/altamirano23a.pdf}, url = {https://proceedings.mlr.press/v202/altamirano23a.html}, abstract = {This paper proposes an online, provably robust, and scalable Bayesian approach for changepoint detection. The resulting algorithm has key advantages over previous work: it provides provable robustness by leveraging the generalised Bayesian perspective, and also addresses the scalability issues of previous attempts. Specifically, the proposed generalised Bayesian formalism leads to conjugate posteriors whose parameters are available in closed form by leveraging diffusion score matching. The resulting algorithm is exact, can be updated through simple algebra, and is more than 10 times faster than its closest competitor.} }
Endnote
%0 Conference Paper %T Robust and Scalable Bayesian Online Changepoint Detection %A Matias Altamirano %A Francois-Xavier Briol %A Jeremias Knoblauch %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-altamirano23a %I PMLR %P 642--663 %U https://proceedings.mlr.press/v202/altamirano23a.html %V 202 %X This paper proposes an online, provably robust, and scalable Bayesian approach for changepoint detection. The resulting algorithm has key advantages over previous work: it provides provable robustness by leveraging the generalised Bayesian perspective, and also addresses the scalability issues of previous attempts. Specifically, the proposed generalised Bayesian formalism leads to conjugate posteriors whose parameters are available in closed form by leveraging diffusion score matching. The resulting algorithm is exact, can be updated through simple algebra, and is more than 10 times faster than its closest competitor.
APA
Altamirano, M., Briol, F. & Knoblauch, J.. (2023). Robust and Scalable Bayesian Online Changepoint Detection. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:642-663 Available from https://proceedings.mlr.press/v202/altamirano23a.html.

Related Material