Title: Sequential change-point detection in panel data models
Authors: Jaewon Huh - University of Texas Southwestern Medical Center (United States) [presenting]
Sangyeol Lee - Seoul National University (Korea, South)
Abstract: A sequential monitoring method is developed for linear panel data models based on the maximum eigenvalue of the empirical covariance matrix. Panel Data models are known to play a major role in analysis of high-dimensional data owing to its similarity to time series models. We pay special attention in extending the change-point detection techniques to the high-dimensional settings with theoretical backup. The performance of the proposed method is evaluated through a simulation study with various parameter settings. Our findings show that the proposed method performs adequately. Moreover, the use of this method is theoretically well equipped.