No label defined (Q4373570)

The authors introduce control charts for monitoring processes of weakly dependent data as well as processes of multivariate quality measurements. Unlike standard methods which require normality for the process distribution, their approach is completely nonparametric. Applying the moving blocks bootstrap method [see \textit{H. R. Künsch}, Ann. Stat. 17, No. 3, 1217-1241 (1989; Zbl 0684.62035); \textit{R. Y. Liu} and \textit{K. Singh}, in R. Le Page et al. (eds.), Exploring the limits of bootstrap. Wiley Ser. Probab. Math. Stat., 225-248 (1992; Zbl 0838.62036)] the authors develop control charts to detect changes of location in data with \(m\)-dependence structure. Their construction of multivariate control charts is based on the concept of data depth and corresponding relative center-outward ranking schemes [\textit{R. Y. Liu} and \textit{K. Singh}, J. Am. Stat. Assoc. 88, No. 421, 252-260 (1993; Zbl 0772.62031)]. In this way, the multivariate observations are reduced to a univariate index which enables easy visualization of the monitoring procedures. Moreover, due to the concept of data depth, these charts are suitable for detecting location shifts and scale increases simultaneously. The authors present some simulation results to demonstrate the validity of the introduced control charts. For the entire collection see [Zbl 0867.00026].