A1133
Title: Asymptotic properties of change point detection in high dimensional settings
Authors: Kento Egashira - Tokyo University of Science (Japan) [presenting]
Kazuyoshi Yata - University of Tsukuba (Japan)
Makoto Aoshima - University of Tsukuba (Japan)
Abstract: Change point detection has received increasing attention in high-dimensional, low-sample-size (HDLSS) settings. While numerous methods have been developed under the assumption of independence across samples, theoretical understanding remains limited when samples exhibit dependency. The asymptotic properties of change point detection are investigated for dependent samples under HDLSS settings, focusing on the method that detects changes in both mean vectors and covariance matrices. Under these conditions, sufficient conditions are derived for the consistency of change point estimators. The theoretical results are further extended to scenarios with multiple change points. Numerical simulations are provided to illustrate the practical performance.
