报告名称：A Composite Likelihood-based Approach for Change-point Detection in Spatio-temporal Process

主办单位：数学与统计学学院

报告专家：邱俊业（YAU Chun Yip）

专家所在单位：香港中文大学统计系

报告时间：2020年7月13日11:00-12:00

报告地点：

Zoom (Meeting ID: 506 231 6553，Password: 123)

Join Zoom Meeting：https://us02web.zoom.us/j/5062316553?pwd=QmpZY0RjYWQ4RXlRVGl4UGY2TWE2UT09

专家简介：邱俊业，统计学博士，毕业于美国哥伦比亚大学，副教授，博士生导师，风险管理科学项目（Risk Management Science Program）的Director。主要研究领域为时间序列、变点分析、经验似然、空间统计与环境统计。在Journal of the Royal Statistical Society - Series B、Journal of the American Statistical Association、Biometrika、Journal of Econometrics等学术期刊上发表SCI论文30余篇。

报告摘要：This paper develops a unified, accurate and computationally efficient method for change-point inference in non-stationary spatio-temporal processes. By modeling a non-stationary spatio-temporal process as a piecewise stationary spatio-temporal process, we consider simultaneous estimation of the number and locations of change-points, and model parameters in each segment. A composite likelihood-based criterion is developed for change-point and parameters estimation. Under the frameworks of increasing domain asymptotics, asymptotic theories including consistency and distribution of the estimators are derived under mild conditions.In contrast to classical results in fixed dimensional time series that the asymptotic error of change-point estimator is O_p(1), exact recovery of true change-points is guaranteed in the spatio-temporal setting. More surprisingly, the consistency of change-point estimation can be achieved without any penalty term in the criterion function. Besides, under a new asymptotic framework in which the time domain is increasing while the spatial sampling domain is fixed, consistency of the number and locations of change-points in piecewise stationary Gaussian processes is also established. A computational efficient pruned dynamic programming algorithm is developed for the challenging criterion optimization problem. Simulation studies and an application to U.S. precipitation data are provided to demonstrate the effectiveness and practicality of the proposed method.

邀请人：刘展