A0412
Title: Some clustering-based change-point detection methods applicable to high dimension, low sample size data
Authors: Angshuman Roy - Indian Institute of Technology Tirupati (India) [presenting]
Trisha Dawn - Texas AM University (United States)
Alokesh Manna - University of Connecticut (United States)
Anil Ghosh - Indian Statistical Institute Kolkata (India)
Abstract: Detecting change points in high-dimensional data is a challenging task, especially when the sample size (i.e., sequence length) is small. Change point detection methods are presented based on clustering, designed to handle such high-dimensional, low-sample-size scenarios effectively. The problem of a single change point is first addressed, introducing methods that leverage k-means clustering with appropriate dissimilarity measures to test for the presence of a change point and estimate its location. These approaches are then extended to detect multiple change points. Through extensive numerical studies and real data analysis, the performance of the proposed methods is evaluated, and those are compared with state-of-the-art techniques.
