Monitoring high-dimensional heteroscedastic processes using rank-based EWMA methods

Zezhong Wang, Rob Goedhart, Inez Maria Zwetsloot

Abstract: Highlights•Efficient in detecting small and sparse changes in high-dimensional processes.•Robustness to correlation among variables and heteroscedasticity over time.•Computational efficiency in processing a large number of variables.•An engineering case study showcases the performance.