|
Activity Number:
|
387
- Software
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 12, 2021 : 12:00 PM to 1:50 PM
|
|
Sponsor:
|
International Statistical Institute
|
|
Abstract #318961
|
|
|
Title:
|
Second-Order Asymptotics of the Empirical Process Under Long-Range Dependence
|
|
Author(s):
|
Marie-Christine Düker* and Annika Betken
|
|
Companies:
|
Cornell University and University of Twente
|
|
Keywords:
|
long-range dependence;
empirical process;
change-point
|
|
Abstract:
|
Due to the slow convergence of the empirical process under long-range dependence, tests for structural changes in the data (based on the asymptotic distribution of the empirical process) are more likely to falsely reject the hypothesis for a small sample size. This talk suggests an alternative approach to determine critical values based on the second order term of the empirical process. The second order term behaves short-range dependent in the sense that it requires the same normalization as the empirical process of i.i.d. data. Empirical results show that the proposed procedure shows favorable performance in comparison to the asymptotic critical values based on the original test statistic as well as in comparison to bootstrapping and subsampling. Theoretical verifications and a data application will also be presented.
|
Authors who are presenting talks have a * after their name.
