A0227
Title: Control charts for monitoring a BINARCH(1) process: Comparisons and applications
Authors: Athanasios Rakitzis - University of Piraeus (Greece) [presenting]
Maria Anastasopoulou - University of the Aegean (Greece)
Abstract: In recent years, control charts for monitoring autocorrelated count data have gained much attention, and there are many scientific works in the area. The most popular time series models of this type are those that assume first-order autocorrelation among the successive counts. The focus is on one of the most popular model in the area, the BINARCH(1) model, which is suitable for modelling first-order autocorrelated binomial counts, i.e., counts that have a bounded support. First, the available control charts in the literature are presented that can be used for the monitoring of this type of process. Then, CUSUM-type control charts are proposed and studied, which have not been studied so far. Using Monte Carlo simulation, the performance of the control charts considered is compared, either for increasing or for decreasing shifts in the process mean level. Recommendations for the most powerful scheme are also given. Finally, their practical implementation is briefly discussed via a real-data example.
