A0852
Title: Monitoring time series with short detection delay
Authors: Tim Kutta - Aarhus University (Denmark) [presenting]
Nina Doernemann - Aarhus University (Denmark)
Abstract: Sequential tests are discussed for a change in the mean of a dependent, Banach space-valued time series. For this purpose, a new class of weighted CUSUM statistics is introduced, which is tailored to the detection of changes shortly after they occur. Unlike current alternatives, which either experience long detection delays or offer short delays only at the very beginning of the monitoring period, the presented approach provides consistently short detection delays anywhere in the monitoring period. This property is highly relevant for modern applications, such as epidemiology and finance, where short delays are crucial and the timing of the change is unpredictable. The theoretical results are based on new Hoelderian invariance principles that are proven under some high-level conditions for Banach space-valued data.
