False Discovery Rate Approach to Dynamic Change Detection

Research output: Journal Publications and ReviewsRGC 21 - Publication in refereed journalpeer-review

View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Original languageEnglish
Article number105224
Number of pages19
Journal / PublicationJournal of Multivariate Analysis
Volume198
Online published1 Aug 2023
Publication statusPublished - Nov 2023

Abstract

In multiple data stream surveillance, the rapid and sequential identification of individuals whose behaviour deviates from the norm has become particularly important. In such applications, the state of a stream can alternate, possibly multiple times, between a null state and an alternative state. To balance the ability to detect two types of changes, that is, a change from the null to the alternative and back to the null, we propose a new multiple testing procedure based on a penalized version of the generalized likelihood ratio test statistics for change detection. The false discovery rate (FDR) at each time point is shown to be controlled under some mild conditions on the dependence structure of data streams. A data-driven approach is developed for selection of the penalization parameter. Its advantage is demonstrated via simulation and a data example. © 2023 Elsevier Inc.

Research Area(s)

  • High-dimensional data streams, Multiple epidemic changes, Multiple testing, Penalized methods, Sequential detection

Citation Format(s)

False Discovery Rate Approach to Dynamic Change Detection. / Du, Lilun; Wen, Mengtao.
In: Journal of Multivariate Analysis, Vol. 198, 105224, 11.2023.

Research output: Journal Publications and ReviewsRGC 21 - Publication in refereed journalpeer-review