A0914
Title: Approximation, estimation and inferential theory for locally stationary functional time series
Authors: Xiucai Ding - UC Davis (United States) [presenting]
Abstract: Some recent results on locally stationary functional time series analysis are reported. First, It is proved that under some mild conditions, every locally stationary functional time series with short-range dependence can be well-approximated by a functional AR process with a diverging number of order which is adaptive to the underlying structures. Second, sieve estimators for the coefficients of the functional AR process, which attains the min-max rates, are provided. Third, inference of these coefficients is conducted by establishing a Gaussian approximation result. Applications include checking the stationarity of the functional time series. A multiplier bootstrap method is proposed for the implementation.