Monitoring Serially Correlated Data by Two CUSUM Charts (Case Study: Numbers of Patients with Covid-19)

Document Type : 18th IIIE conference selected papers


Department of Industrial Engineering, University of Kurdistan



Statistical process control charts have many applications in manufacturing industries, environmental monitoring and improvement, disease surveillance, and more. Statistical process control charts usually are designed for cases when process observations are independent at different observation times. However, serial data correlation almost always exists in sequential data. Thus, it is important to develop control charts specially for monitoring serially correlated data. On the other hand, one of the most important cases today is the Covid-19 epidemic, and it has been proven that any infected person can infect other people whose symptoms appear a few days later. The main purpose of this paper, is to monitor the condition of covid-19 patients during specified time using serially data. To this aim, we use two new CUSUM charts to monitor the number of patients with Covid-19, which runs for three countries include Iran, Japan and Italy. The results displayed separately for each country and explained with appropriate tools. Meanwhile, a sensitivity analysis on important factors is performed and similar results are obtained and those two control charts are compared.