A0312
Title: $L_2-L_\infty$ inference of breaks for high dimensional time series
Authors: Likai Chen - Washington University in Saint Louis (United States) [presenting]
Jiaqi Li - Washington University in Saint Louis (United States)
Weining Wang - City U of London (United Kingdom)
Wei Biao Wu - University of Chicago (United States)
Abstract: A new method is proposed for multiple change points detection of high-dimensional time series. The proposed approach targets dense or clustered cross-sectional signals. An $L_2$-aggregated statistics is adopted within each rectangular window and $L_\infty$ aggregation is applied for all the windows. On the theory front, we develop an asymptotic theory concerning the limiting distributions of the change-point test statistics under both the null and alternatives and the consistency of the estimated break dates. The core of our theory is to extend the high-dimensional Gaussian approximation theorem for $L_2$-based statistics for dependent data. Weakly temporal and cross-sectional dependences can be allowed. Simulations show the power enhancement in the presence of dense or clustered signals relative to the maximum statistics.
