Selective review of offline change point detection methods

• Published in: Signal processing Vol. 167; p. 107299
• Main Authors: Truong, Charles, Oudre, Laurent, Vayatis, Nicolas
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.02.2020; Elsevier
• Subjects: Change point detection; Segmentation; Statistical signal processing; Statistics; Statistics Theory; Segmentation; Statistical signal processing; Change point detection
• ISSN: 0165-1684; 1872-7557
• DOI: 10.1016/j.sigpro.2019.107299

•A structured and didactic review of more than 140 articles related to offline change point detection. Thanks to the methodological framework proposed in this survey, all methods are presented as the combination of three functional blocks, which facilitates comparison between the different approaches.•The survey provides details on mathematical as well as algorithmic aspects such as complexity, asymptotic consistency, estimation of the number of changes, calibration, etc.•The review is linked to a Python package that includes most of the pre- sented methods, and allows the user to perform experiments and bench- marks. This article presents a selective survey of algorithms for the offline detection of multiple change points in multivariate time series. A general yet structuring methodological strategy is adopted to organize this vast body of work. More precisely, detection algorithms considered in this review are characterized by three elements: a cost function, a search method and a constraint on the number of changes. Each of those elements is described, reviewed and discussed separately. Implementations of the main algorithms described in this article are provided within a Python package called ruptures.