报告题目：On Value-at-Risk Based Queueing-Game
报告人: 王金亭 中央财经大学管理科学与工程学院
报告时间：2024年10月27日（星期日）上午10:00
报告地点：仙林校区教2-327会议室
主办单位：565net必赢客户端565net必赢客户端
邀请人：步琪慧
报告内容：In this talk, we focus on the economics of queueing systems with strategic non-risk-neutral customers, emphasizing the value-at-risk (VaR) framework. It is distinct from the moment-based utility structure of Naor's model by introducing the maximum loss customers can bear with a certain level of confidence. Such systems are ubiquitous in various domains involving human customers, such as healthcare or security check queues, wherein risk-based performance measures take precedence and hold greater significance. We consider both homogeneous and heterogeneous cases of risk preferences under different information levels. Our findings demonstrate that optimal strategies in observable queueing systems with non-risk-neutral customers exhibit threshold-type behaviors that differ significantly from those observed in risk-neutral customer settings. For unobservable queueing systems, we derive an equilibrium joining probability for homogeneous risk preferences and a multi-dimensional equilibrium joining probability for heterogeneous risk preferences. Interestingly, under the VaR criterion, there is an indifferent action region in the observable queueing model. Customers with risk preferences falling within this region will adopt the same joining strategy. And, in the unobservable scenario, there is an indifference curve for risk preference. Customers on the same indifference curve have the same level of risk preference. Furthermore, we provide a conversion formula that facilitates comparison between two risk preferences with different confidence levels. When exploring social welfare, we use conditional value at risk (CVaR) to characterize the customer's excessive losses that lead to potential negative social utility. In observable systems with homogeneous risk preferences, the socially optimal risk preference falls within a “ribbon” region, while in unobservable systems, it forms a curve. Overall, this research sheds light on the interplay between VaR-based queueing economics, risk preferences, and strategic customer behavior.
报告人简介：王金亭，男，博士，中央财经大学二级教授、龙马学者特聘教授、博士生导师、中央财经大学第九届学术委员会委员、管理科学与工程学院学术委员会主任、学科带头人，国家级一流本科专业“管理科学”专业建设负责人、中央财经大学“智能管理”交叉学科建设项目首席科学家。兼任中国运筹学会常务理事、中国运筹学会可靠性分会理事长等社团职务。2004年获中国运筹学会科学技术奖之青年科技奖（原中国运筹学会青年运筹奖），2006年获教育部霍英东教育基金会第十届高等院校青年教师奖，2011年入选教育部新世纪优秀人才计划，2016年获得第八届安徽省自然科学优秀论文一等奖，2018年获詹天佑铁道科学技术奖专项科技奖，入选2021、2023年度Elsevier“高被引学者”榜单；多次入选斯坦福大学发布的全球前2%顶尖科学家排名榜单。从事随机运筹学、运作管理、排队经济学、供应链优化与管理等方面的研究与教学工作。主持完成7项国家自然科学基金项目，参与承担国家自然科学基金重大项目“城市物流”等20余项国家级/省部级课题。发表学术论文160余篇（120余篇SCI及30余篇SSCI论文），研究成果发表在包括Operations Research、Manufacturing & Service Operations Management、Production and Operations Management、European Journal of Operational Research, International Journal of Production Economics、International Journal of Production Research, Queueing Systems, Reliability Engineering & System Safety、IEEE Transactions多个汇刊在内的国际学术期刊上。2016年出版专著《排队博弈论基础》（科学出版社，独著）。2022年获第15届EAI无线互联网国际会议最佳论文奖，及全国供应链与运营管理学术年会最佳论文一等奖等学术奖励。2023年获评为中央财经大学优秀研究生指导教师。2024年10月获中国运筹学会科学技术奖运筹研究奖。