Monitoring the coefficient of variation using variable sampling interval CUSUM control charts. Issue 3 (11th February 2021)

• Record Type: Journal Article
• Authors: Tran, P. H.; Heuchenne, C.
• Is Part Of: Journal of statistical computation and simulation. Volume 91:Issue 3(2021)
• Page Start: 501
• Page End: 521
• Publication Date: 2021-02-11
• Subjects: VSI -- CUSUM -- coefficient of variation -- Markov chain -- anomaly detection; 60J10 -- 62P30; Mathematical statistics -- Data processing -- Periodicals; Digital computer simulation -- Periodicals; 519.5028505
• DOI: 10.1080/00949655.2020.1819278
• Languages: English
• ISSNs: 0094-9655

Abstract

Abstract : In monitoring the coefficient of variation (CV), the variable sampling interval (VSI) type charts are a useful adaptive strategy that has been recently combined with various types of charts, including exponentially weighted moving average (EWMA) and variable sample size (VSS) charts. These combinations lead to desirable control charts with an impressive performance in detecting process shifts or anomalies. However, in these studies, the authors asserted that adaptive type charts are not competitive to monitor downward shifts in the CV. Therefore, they willingly omitted the decrease in process shifts. In this paper, we suggest overcoming this downside by designing two one-sided VSI CUSUM control charts for tracking the CV. The performance of the proposed charts is evaluated by the average time to signal (ATS) criterion. The numerical results show that our proposed adaptive type charts are efficient charts to monitor downward shifts in the CV.