l2 INFERENCE FOR CHANGE POINTS IN HIGH-DIMENSIONAL TIMESERIES VIA A TWO-WAY MOSUM

• 成果类型: Article
• 署名作者: Li, Jiaqi; Chen, Likai; Wang, Weining; Wu, Wei biao
• 署名单位: Washington University (WUSTL); University of Groningen; University of Chicago
• 刊物名称: ANNALS OF STATISTICS
• ISSN/ISSBN: 0090-5364
• DOI: 10.1214/24-AOS2360
• 发表日期: 2024
• 页码: 602-627
• 关键词: binary segmentation; time-series; panel-data; bootstrap; cluster; models; tests

摘要

We propose an inference method for detecting multiple change points inhigh-dimensional time series, targeting dense or spatially clustered signals.Our method aggregates moving sum (MOSUM) statistics cross-sectionallyby an2-norm and maximizes them over time. We further introduce anovel Two-Way MOSUM, which utilizes spatial-temporal moving regionsto search for breaks, with the added advantage of enhancing testing powerwhen breaks occur in only a few groups. The limiting distribution of an2-aggregated statistic is established for testing break existence by extending ahigh-dimensional Gaussian approximation theorem to spatial-temporal non-stationary processes. Simulation studies exhibit promising performance ofour test in detecting nonsparse weak signals. Two applications on equity re-turns and COVID-19 cases in the United States show the real-world relevanceof our algorithms. The R package L2hdchange is available on CRAN.