Home > Research > Publications & Outputs > Consistency of a range of penalised cost approa...

Research

Associated organisational units

Electronic data

  • ejs-revision

    Accepted author manuscript, 298 KB, PDF document

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

Links

Text available via DOI:

Keywords

View graph of relations

Consistency of a range of penalised cost approaches for detecting multiple changepoints

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Published

Standard

Consistency of a range of penalised cost approaches for detecting multiple changepoints. / Zheng, Chao; Eckley, Idris; Fearnhead, Paul.
In: Electronic Journal of Statistics, Vol. 16, No. 2, 30.08.2022, p. 4497-4546.

Research output: Contribution to Journal/Magazine › Journal article › peer-review

Harvard

Zheng, C, Eckley, I & Fearnhead, P 2022, 'Consistency of a range of penalised cost approaches for detecting multiple changepoints', Electronic Journal of Statistics, vol. 16, no. 2, pp. 4497-4546. https://doi.org/10.1214/22-EJS2048

APA

Zheng, C., Eckley, I., & Fearnhead, P. (2022). Consistency of a range of penalised cost approaches for detecting multiple changepoints. Electronic Journal of Statistics, 16(2), 4497-4546. https://doi.org/10.1214/22-EJS2048

Vancouver

Zheng C, Eckley I, Fearnhead P. Consistency of a range of penalised cost approaches for detecting multiple changepoints. Electronic Journal of Statistics. 2022 Aug 30;16(2):4497-4546. doi: 10.1214/22-EJS2048

Author

Zheng, Chao ; Eckley, Idris ; Fearnhead, Paul. / Consistency of a range of penalised cost approaches for detecting multiple changepoints. In: Electronic Journal of Statistics. 2022 ; Vol. 16, No. 2. pp. 4497-4546.

Bibtex

@article{22dab9f368d3473283984a609df85cb1,
title = "Consistency of a range of penalised cost approaches for detecting multiple changepoints",
abstract = "A common approach to detect multiple changepoints is to min- imise a measure of data fit plus a penalty that is linear in the number of changepoints. This paper shows that the general finite sample behaviour of such a method can be related to its behaviour when analysing data with either none or one changepoint. This property results in simpler conditions for verifying whether the method will consistently estimate the number and locations of the changepoints. We apply and demonstrate the useful- ness of these simple conditions for a range of changepoint problems. Our new results include a weaker requirement on the choice of penalty to have consistency in a change-in-slope model; and the first results for the accuracy of recently-proposed methods for detecting spikes.",
keywords = "Change-point detection, consistency, local region conditions, changes in slope, spike plus exponential decay",
author = "Chao Zheng and Idris Eckley and Paul Fearnhead",
year = "2022",
month = aug,
day = "30",
doi = "10.1214/22-EJS2048",
language = "English",
volume = "16",
pages = "4497--4546",
journal = "Electronic Journal of Statistics",
issn = "1935-7524",
publisher = "Institute of Mathematical Statistics",
number = "2",

}

RIS

TY - JOUR

T1 - Consistency of a range of penalised cost approaches for detecting multiple changepoints

AU - Zheng, Chao

AU - Eckley, Idris

AU - Fearnhead, Paul

PY - 2022/8/30

Y1 - 2022/8/30

N2 - A common approach to detect multiple changepoints is to min- imise a measure of data fit plus a penalty that is linear in the number of changepoints. This paper shows that the general finite sample behaviour of such a method can be related to its behaviour when analysing data with either none or one changepoint. This property results in simpler conditions for verifying whether the method will consistently estimate the number and locations of the changepoints. We apply and demonstrate the useful- ness of these simple conditions for a range of changepoint problems. Our new results include a weaker requirement on the choice of penalty to have consistency in a change-in-slope model; and the first results for the accuracy of recently-proposed methods for detecting spikes.

AB - A common approach to detect multiple changepoints is to min- imise a measure of data fit plus a penalty that is linear in the number of changepoints. This paper shows that the general finite sample behaviour of such a method can be related to its behaviour when analysing data with either none or one changepoint. This property results in simpler conditions for verifying whether the method will consistently estimate the number and locations of the changepoints. We apply and demonstrate the useful- ness of these simple conditions for a range of changepoint problems. Our new results include a weaker requirement on the choice of penalty to have consistency in a change-in-slope model; and the first results for the accuracy of recently-proposed methods for detecting spikes.

KW - Change-point detection

KW - consistency

KW - local region conditions

KW - changes in slope

KW - spike plus exponential decay

U2 - 10.1214/22-EJS2048

DO - 10.1214/22-EJS2048

M3 - Journal article

VL - 16

SP - 4497

EP - 4546

JO - Electronic Journal of Statistics

JF - Electronic Journal of Statistics

SN - 1935-7524

IS - 2

ER -