Adaptive Change Point Inference for High Dimensional Time Series with Temporal Dependence

Xiaoyi Wang, Jixuan Liu, Long Feng, Bejing Normal University, Nankai University · 2025

Read Adaptive Change Point Inference for High Dimensional Time Series with Temporal Dependence by Xiaoyi Wang, Jixuan Liu, Long Feng, Bejing Normal University, Nankai University (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-01487.txt.

Description

Xiaoyi Wang, Jixuan Liu, and Long Feng develop tests for detecting change points, moments when a time series' behavior shifts, in data with very many variables tracked over time. Their starting test uses a maximum L2-norm statistic, which performs strongly when the change is spread densely across many variables.

Sparse changes, concentrated in a few variables, are better caught by other statistics based on the maximum coordinate. The authors prove their statistic is asymptotically independent of two such existing statistics, which lets them merge all three with the Cauchy combination method into one adaptive procedure. The combined test stays robust whether the change is dense or sparse, and simulations plus a real-data analysis confirm its advantage in high-dimensional settings.

Catalog record

CREATORXiaoyi Wang, Jixuan Liu, Long Feng, Bejing Normal University, Nankai University
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE207 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-01487

Browse the full library · About the Tower of Babel