报告题目:Some New Real-time Monitoring Schemes for Gumbel’s Bivariate Exponential Time between the Events
报告时间:2025年5月29日下午15:00
报告地点:南湖校区老图书馆四楼会议室
主办单位:数学与统计学院
报告人:张久军
报告人简介:张久军,辽宁大学数学与统计学院教授,博士研究生导师,统计与数据科学系主任。中国现场统计研究会生存分析分会理事,全国工业统计学教学研究会多元统计分析分会理事,辽宁省数学会理事。辽宁省教学名师,辽宁大学中青年骨干教师,本科优秀主讲教师,第九届校学术委员会委员。辽宁省“百千万人才工程”百人层次人选,“沈阳市高级人才-领军人才”。主持、参与国家自然科学基金,辽宁省自然科学基金等多项,发表论文50余篇。
摘要:Monitoring the vector of times between multiple events is essential in a high-quality process such as healthcare operations. To this end, the multivariate time between events (TBE) process monitoring schemes are regularly used as one of the most straightforward and appealing visual tools. The existing literature on multivariate TBE schemes focuses almost exclusively on using complete information availed in vector-based TBE data, often making delayed monitoring as it requires observing the complete set of time values in a vector-valued observation. To address this issue, we recommend monitoring the minimum time value of vector TBE data to reach decisions faster and more efficiently. We introduce several new real-time exponentially weighted moving average (EWMA) schemes for monitoring Gumbel's bivariate exponential TBE processes. We compare them with existing schemes using fully observed vector-based schemes. A Markov chain method is developed to compute the average time to signal (ATS), and the optimal parameters are found. Finally, three real-life examples are used to illustrate the implementation of the proposed schemes.