EcoSta 2022: Start Registration
View Submission - EcoSta2022
A0152
Title: Multiple change-point detection for functional data (virtual) Authors:  Jeng-Min Chiou - Academia Sinica (Taiwan) [presenting]
Abstract: Detecting abrupt structural changes in a data sequence is interesting in many applications. Multiple change-point detection of a functional data sequence is discussed through two recent approaches: dynamic segmentation and greedy segmentation. They comprise the detection and the significance testing procedures under the least-squares segmentation criterion intending to identify the locations and the number of change points. We discuss the asymptotic properties of the methods and explore their finite sample performance through simulation studies and data applications.