Gaussian Process Subset Scanning for Anomalous Pattern Detection in Non-iid Data

William Herlands, Edward McFowland, Andrew Wilson, Daniel Neill
Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, PMLR 84:425-434, 2018.

Abstract

Identifying anomalous patterns in real-world data is essential for understanding where, when, and how systems deviate from their expected dynamics. Yet methods that separately consider the anomalousness of each individual data point have low detection power for subtle, emerging irregularities. Additionally, recent detection techniques based on subset scanning make strong independence assumptions and suffer degraded performance in correlated data. We introduce methods for identifying anomalous patterns in non-iid data by combining Gaussian processes with novel log-likelihood ratio statistic and subset scanning techniques. Our approaches are powerful, interpretable, and can integrate information across multiple streams. We illustrate their performance on numeric simulations and three open source spatiotemporal datasets of opioid overdose deaths, 311 calls, and storm reports.

Cite this Paper

BibTeX
@InProceedings{pmlr-v84-herlands18a, title = {Gaussian Process Subset Scanning for Anomalous Pattern Detection in Non-iid Data}, author = {Herlands, William and McFowland, Edward and Wilson, Andrew and Neill, Daniel}, booktitle = {Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics}, pages = {425--434}, 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/herlands18a/herlands18a.pdf}, url = {https://proceedings.mlr.press/v84/herlands18a.html}, abstract = {Identifying anomalous patterns in real-world data is essential for understanding where, when, and how systems deviate from their expected dynamics. Yet methods that separately consider the anomalousness of each individual data point have low detection power for subtle, emerging irregularities. Additionally, recent detection techniques based on subset scanning make strong independence assumptions and suffer degraded performance in correlated data. We introduce methods for identifying anomalous patterns in non-iid data by combining Gaussian processes with novel log-likelihood ratio statistic and subset scanning techniques. Our approaches are powerful, interpretable, and can integrate information across multiple streams. We illustrate their performance on numeric simulations and three open source spatiotemporal datasets of opioid overdose deaths, 311 calls, and storm reports.} }
Endnote
%0 Conference Paper %T Gaussian Process Subset Scanning for Anomalous Pattern Detection in Non-iid Data %A William Herlands %A Edward McFowland %A Andrew Wilson %A Daniel Neill %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-herlands18a %I PMLR %P 425--434 %U https://proceedings.mlr.press/v84/herlands18a.html %V 84 %X Identifying anomalous patterns in real-world data is essential for understanding where, when, and how systems deviate from their expected dynamics. Yet methods that separately consider the anomalousness of each individual data point have low detection power for subtle, emerging irregularities. Additionally, recent detection techniques based on subset scanning make strong independence assumptions and suffer degraded performance in correlated data. We introduce methods for identifying anomalous patterns in non-iid data by combining Gaussian processes with novel log-likelihood ratio statistic and subset scanning techniques. Our approaches are powerful, interpretable, and can integrate information across multiple streams. We illustrate their performance on numeric simulations and three open source spatiotemporal datasets of opioid overdose deaths, 311 calls, and storm reports.
APA
Herlands, W., McFowland, E., Wilson, A. & Neill, D.. (2018). Gaussian Process Subset Scanning for Anomalous Pattern Detection in Non-iid Data. Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 84:425-434 Available from https://proceedings.mlr.press/v84/herlands18a.html.

Related Material