2013/06/20 by David S. Matteson, Matteson, David S., Nicholas A. James +1 · 8 citations
Biochemistry, Genetics and Molecular Biology · Mathematics · #FOS: Computer and information sciences #Genetic Mapping and Diversity in Plants and Animals #Methodology (stat.ME) #Statistical Methods and Bayesian Inference #Statistical Methods and Inference
paper · pdf · doi:10.48550/arxiv.1306.4933
openalex publication_date 2013/06/20 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Change point analysis has applications in a wide variety of fields. The\ngeneral problem concerns the inference of a change in distribution for a set of\ntime-ordered observations. Sequential detection is an online version in which\nnew data is continually arriving and is analyzed adaptively. We are concerned\nwith the related, but distinct, offline version, in which retrospective\nanalysis of an entire sequence is performed. For a set of multivariate\nobservations of arbitrary dimension, we consider nonparametric estimation of\nboth the number of change points and the positions at which they occur. We do\nnot make any assumptions regarding the nature of the change in distribution or\nany distribution assumptions beyond the existence of the alpha-th absolute\nmoment, for some alpha in (0,2). Estimation is based on hierarchical clustering\nand we propose both divisive and agglomerative algorithms. The divisive method\nis shown to provide consistent estimates of both the number and location of\nchange points under standard regularity assumptions. We compare the proposed\napproach with competing methods in a simulation study. Methods from cluster\nanalysis are applied to assess performance and to allow simple comparisons of\nlocation estimates, even when the estimated number differs. We conclude with\napplications in genetics, finance and spatio-temporal analysis.\n