Home > Research > Publications & Outputs > A Log-Linear Non-Parametric Online Changepoint ...

Electronic data

  • NP-FOCuS

    Accepted author manuscript, 3.42 MB, PDF document

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

Links

Text available via DOI:

View graph of relations

A Log-Linear Non-Parametric Online Changepoint Detection Algorithm based on Functional Pruning

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

A Log-Linear Non-Parametric Online Changepoint Detection Algorithm based on Functional Pruning. / Romano, Gaetano; Eckley, Idris; Fearnhead, Paul.
In: IEEE Transactions on Signal Processing, Vol. 72, 31.01.2024, p. 594 - 606.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

APA

Vancouver

Romano G, Eckley I, Fearnhead P. A Log-Linear Non-Parametric Online Changepoint Detection Algorithm based on Functional Pruning. IEEE Transactions on Signal Processing. 2024 Jan 31;72:594 - 606. Epub 2023 Dec 19. doi: 10.1109/TSP.2023.3343550

Author

Bibtex

@article{6ffc9c17bef247eb860e57d925a7c778,
title = "A Log-Linear Non-Parametric Online Changepoint Detection Algorithm based on Functional Pruning",
abstract = "Online changepoint detection aims to detect anomalies and changes in real time within high frequency data streams, sometimes with limited available computational resources. This is an important task that is rooted in many real-world applications including, but not limited to, cybersecurity, medicine and astrophysics. While fast and efficient online algorithms have been recently introduced, these rely on parametric assumptions which are often violated in practical applications. Motivated by data streams from the telecommunications sector, we build a flexible nonparametric approach to detect a change in the distribution of a sequence. Our procedure, NP-FOCuS, builds a sequential likelihood ratio test for a change, across at a set of points, of the empirical cumulative density function of our data. This is achieved by keeping track of the number of observations above or below those points. Thanks to functional pruning ideas, NP-FOCuS has a computational cost that is log-linear in the number of observations and is suitable for high-frequency data streams. In terms of detection power, NP-FOCuS is seen to outperform current nonparametric online changepoint techniques in a variety of settings. We demonstrate the utility of the procedure on both simulated and real data.",
keywords = "Changepoint, anomaly detection, non-parametric, online, real-time analysis, telecommunications",
author = "Gaetano Romano and Idris Eckley and Paul Fearnhead",
year = "2024",
month = jan,
day = "31",
doi = "10.1109/TSP.2023.3343550",
language = "English",
volume = "72",
pages = "594 -- 606",
journal = "IEEE Transactions on Signal Processing",
issn = "1053-587X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

RIS

TY - JOUR

T1 - A Log-Linear Non-Parametric Online Changepoint Detection Algorithm based on Functional Pruning

AU - Romano, Gaetano

AU - Eckley, Idris

AU - Fearnhead, Paul

PY - 2024/1/31

Y1 - 2024/1/31

N2 - Online changepoint detection aims to detect anomalies and changes in real time within high frequency data streams, sometimes with limited available computational resources. This is an important task that is rooted in many real-world applications including, but not limited to, cybersecurity, medicine and astrophysics. While fast and efficient online algorithms have been recently introduced, these rely on parametric assumptions which are often violated in practical applications. Motivated by data streams from the telecommunications sector, we build a flexible nonparametric approach to detect a change in the distribution of a sequence. Our procedure, NP-FOCuS, builds a sequential likelihood ratio test for a change, across at a set of points, of the empirical cumulative density function of our data. This is achieved by keeping track of the number of observations above or below those points. Thanks to functional pruning ideas, NP-FOCuS has a computational cost that is log-linear in the number of observations and is suitable for high-frequency data streams. In terms of detection power, NP-FOCuS is seen to outperform current nonparametric online changepoint techniques in a variety of settings. We demonstrate the utility of the procedure on both simulated and real data.

AB - Online changepoint detection aims to detect anomalies and changes in real time within high frequency data streams, sometimes with limited available computational resources. This is an important task that is rooted in many real-world applications including, but not limited to, cybersecurity, medicine and astrophysics. While fast and efficient online algorithms have been recently introduced, these rely on parametric assumptions which are often violated in practical applications. Motivated by data streams from the telecommunications sector, we build a flexible nonparametric approach to detect a change in the distribution of a sequence. Our procedure, NP-FOCuS, builds a sequential likelihood ratio test for a change, across at a set of points, of the empirical cumulative density function of our data. This is achieved by keeping track of the number of observations above or below those points. Thanks to functional pruning ideas, NP-FOCuS has a computational cost that is log-linear in the number of observations and is suitable for high-frequency data streams. In terms of detection power, NP-FOCuS is seen to outperform current nonparametric online changepoint techniques in a variety of settings. We demonstrate the utility of the procedure on both simulated and real data.

KW - Changepoint

KW - anomaly detection

KW - non-parametric

KW - online

KW - real-time analysis

KW - telecommunications

U2 - 10.1109/TSP.2023.3343550

DO - 10.1109/TSP.2023.3343550

M3 - Journal article

VL - 72

SP - 594

EP - 606

JO - IEEE Transactions on Signal Processing

JF - IEEE Transactions on Signal Processing

SN - 1053-587X

ER -