Home > Research > Publications & Outputs > Scalable change-point and anomaly detection in ...

Electronic data

  • AOAS2010-016R2A0

    Rights statement: The final, definitive version of this article has been published in the Journal, Annals of Applied Statistics, 16 (2), pp 721-743 2022, © 2022 Cambridge University Press.

    Accepted author manuscript, 3.91 MB, PDF document

    Available under license: CC BY-NC: Creative Commons Attribution-NonCommercial 4.0 International License

Links

Text available via DOI:

View graph of relations

Scalable change-point and anomaly detection in cross-correlated data with an application to condition monitoring

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

Scalable change-point and anomaly detection in cross-correlated data with an application to condition monitoring. / Tveten, Martin; Eckley, Idris; Fearnhead, Paul.
In: Annals of Applied Statistics, Vol. 16, No. 2, 30.06.2022, p. 721-743.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

APA

Vancouver

Tveten M, Eckley I, Fearnhead P. Scalable change-point and anomaly detection in cross-correlated data with an application to condition monitoring. Annals of Applied Statistics. 2022 Jun 30;16(2):721-743. doi: 10.1214/21-AOAS1508

Author

Bibtex

@article{2265749130ea45a9b28d564768744d33,
title = "Scalable change-point and anomaly detection in cross-correlated data with an application to condition monitoring",
abstract = "Motivated by a condition monitoring application arising from subsea engineering, we derive a novel, scalable approach to detecting anomalous mean structure in a subset of correlated multivariate time series. Given the need to analyse such series efficiently, we explore a computationally efficient approximation of the maximum likelihood solution to the resulting modelling framework and develop a new dynamic programming algorithm for solving the resulting binary quadratic programme when the precision matrix of the time series at any given time point is banded. Through a comprehensive simulation study we show that the resulting methods perform favorably compared to competing methods, both in the anomaly and change detection settings, even when the sparsity structure of the precision matrix estimate is misspecified. We also demonstrate its ability to correctly detect faulty time periods of a pump within the motivating application.",
keywords = "Anomaly, Binary Quadratic Programme, Changepoints, Cross-correlation, Outliers",
author = "Martin Tveten and Idris Eckley and Paul Fearnhead",
note = "The final, definitive version of this article has been published in the Journal, Annals of Applied Statistics, 16 (2), pp 721-743 2022, {\textcopyright} 2022 Cambridge University Press.",
year = "2022",
month = jun,
day = "30",
doi = "10.1214/21-AOAS1508",
language = "English",
volume = "16",
pages = "721--743",
journal = "Annals of Applied Statistics",
issn = "1932-6157",
publisher = "Institute of Mathematical Statistics",
number = "2",

}

RIS

TY - JOUR

T1 - Scalable change-point and anomaly detection in cross-correlated data with an application to condition monitoring

AU - Tveten, Martin

AU - Eckley, Idris

AU - Fearnhead, Paul

N1 - The final, definitive version of this article has been published in the Journal, Annals of Applied Statistics, 16 (2), pp 721-743 2022, © 2022 Cambridge University Press.

PY - 2022/6/30

Y1 - 2022/6/30

N2 - Motivated by a condition monitoring application arising from subsea engineering, we derive a novel, scalable approach to detecting anomalous mean structure in a subset of correlated multivariate time series. Given the need to analyse such series efficiently, we explore a computationally efficient approximation of the maximum likelihood solution to the resulting modelling framework and develop a new dynamic programming algorithm for solving the resulting binary quadratic programme when the precision matrix of the time series at any given time point is banded. Through a comprehensive simulation study we show that the resulting methods perform favorably compared to competing methods, both in the anomaly and change detection settings, even when the sparsity structure of the precision matrix estimate is misspecified. We also demonstrate its ability to correctly detect faulty time periods of a pump within the motivating application.

AB - Motivated by a condition monitoring application arising from subsea engineering, we derive a novel, scalable approach to detecting anomalous mean structure in a subset of correlated multivariate time series. Given the need to analyse such series efficiently, we explore a computationally efficient approximation of the maximum likelihood solution to the resulting modelling framework and develop a new dynamic programming algorithm for solving the resulting binary quadratic programme when the precision matrix of the time series at any given time point is banded. Through a comprehensive simulation study we show that the resulting methods perform favorably compared to competing methods, both in the anomaly and change detection settings, even when the sparsity structure of the precision matrix estimate is misspecified. We also demonstrate its ability to correctly detect faulty time periods of a pump within the motivating application.

KW - Anomaly

KW - Binary Quadratic Programme

KW - Changepoints

KW - Cross-correlation

KW - Outliers

U2 - 10.1214/21-AOAS1508

DO - 10.1214/21-AOAS1508

M3 - Journal article

VL - 16

SP - 721

EP - 743

JO - Annals of Applied Statistics

JF - Annals of Applied Statistics

SN - 1932-6157

IS - 2

ER -