EcoSta 2019: Start Registration
View Submission - EcoSta2019
Title: A self-normalized approach to sequential change-point detection for time series Authors:  Wai Leong Ng - The Hang Seng University of Hong Kong (Hong Kong) [presenting]
Chun Yip Yau - Chinese University of Hong Kong (Hong Kong)
NH Chan - The Chinese University of Hong Kong (Hong Kong)
Abstract: A self-normalization sequential change-point detection method for time series is proposed. In testing for parameter changes, most of the traditional sequential monitoring tests utilize a CUSUM-based test statistic, which involves a long-run variance estimator. However, the commonly used long-run variance estimators require the choice of bandwidth parameter which could be sensitive to the performance. Moreover, the traditional tests usually suffer from severe size distortion due to the slow convergence rate to the limiting distribution in the early monitoring stage. A self-normalization approach is implemented to tackle these issues. We establish null asymptotic and the consistency of the proposed sequential change-point test under general regularity conditions. Simulation experiments and empirical application to railway bearing temperature data are conducted for illustrations.