Robust quickest correlation change detection in high-dimensional random vectors

Detecting changes in high-dimensional vectors presents significant challenges, especially when the postchange distribution is unknown and time-varying. This article introduces a novel robust algorithm for correlation change detection in high-dimensional data. The algorithm is based on the maximum ma...

Full description

Saved in:
Bibliographic Details
Published inSequential analysis Vol. 44; no. 3; pp. 273 - 292
Main Authors Alghamdi, Assma, Banerjee, Taposh, Rajgopal, Jayant
Format Journal Article
LanguageEnglish
Published Taylor & Francis 03.07.2025
Subjects
High-dimensional vectors
maximum magnitude sample correlation
robust correlation change detection
summary statistics
Online AccessGet full text

Cover

More Information
Summary:Detecting changes in high-dimensional vectors presents significant challenges, especially when the postchange distribution is unknown and time-varying. This article introduces a novel robust algorithm for correlation change detection in high-dimensional data. The algorithm is based on the maximum magnitude correlation coefficient and its approximate asymptotic density. The coefficient captures the level of correlation present in the data. The proposed approach is robust because it can help detect a change in correlation level from some known level to unknown, time-varying levels. The proposed test is also computationally efficient and valid for a broad class of data distributions. The effectiveness of the proposed algorithm is demonstrated on simulated data.
ISSN:0747-4946
1532-4176
DOI:10.1080/07474946.2025.2485172