A0582
Title: Inference for quantile change points in high-dimensional time series
Authors: Likai Chen - Washington University in Saint Louis (United States)
Jiaqi Li - University of Chicago (United States)
Mengyu Xu - University of Central Florida (United States) [presenting]
Abstract: Change-point detection methods based on quantiles can effectively detect changes in extreme values. A novel change-point detection scheme that utilizes fixed quantiles of moving sums from high-dimensional time series data is proposed. The approach employs a moving sum (MOSUM) test statistic aggregating the component series by the $\ell^{\infty}$ norm. The asymptotic properties of the proposed test statistic are investigated in the context of weak temporal-dependent high-dimensional time series while also allowing for strong and weak cross-sectional dependence. The analysis relies on a powerful uniform Bahadur representation result. Specifically, the existing uniform Bahadur representation is extended to the high-dimensional setting for dependent data. A simulation study demonstrates the effectiveness of the approach.
