CMStatistics 2017: Start Registration
View Submission - CMStatistics
Title: Robust change-point estimation in the presence of long-range dependence Authors:  Annika Betken - Ruhr-Universitat Bochum (Germany) [presenting]
Abstract: An estimator is considered for the location of a shift in the mean of long-range dependent sequences. The estimation is based on Wilcoxon-type statistics. Consistency and an optimal rate of convergence are established for constant shift heights and and under certain local changes, i.e. for shift heights decreasing to 0 with a certain rate. If suitably standardized, the estimator converges in distribution to a functional of a fractional Brownian motion process. Its asymptotic behavior as well as its finite sample performance are compared to corresponding properties of a CUSUM-type change-point estimator.