报告题目：Optimal QueueLength Information Disclosure When Service Quality is Uncertain

报告人：郭朋飞（香港理工大学）

报告人简介：郭朋飞现为香港理工大学物流与航运学系教授，他本科毕业于西安交通大学，硕士毕业于上海交通大学，博士毕业于美国杜克大学。他的研究兴趣和成果主要集中在排队经济学领域。他同时对库存管理，供应链管理和医疗卫生系统的运营问题都有广泛的涉猎和研究。郭朋飞教授在ManagementScience, Operations Research, M&SOM and POM等管理科学/运筹学顶级刊物上发表30余篇论文，目前是UTD期刊之一POM的高级编辑（SeniorEditor）。

报告内容摘要：Consider a single-server service system with uncertain qualitylevel (which is assumed to be binary). Both the server and the customers knowthe distribution of quality levels and are engaged in the following two-stagegame. In the first stage, the server commits to a strategy (possibly mixed)that, given a realized quality level, states whether or not the queue lengthwill be revealed to customers upon their arrival. In the second stage, qualityis realized and the server's corresponding queue-disclosure action is observedby the customers, who then update their belief on the service quality and decidewhether or not to join the service system. The server's decision problem is tofind an optimal commitment strategy to maximize his expected effective arrivalrate, anticipating the customers' equilibrium queueing behavior toward hisqueue-disclosure strategy. We reformulate the server's decision problem aslooking for the best resulting posterior distribution on service quality. Thisreformulation yields useful insights into when and why a commitment strategyhelps. In particular, we demonstrate that the maximal expected effectivearrival rate, as a function of the prior, can be graphed as the upper envelopeof all convex combinations of points on the effective arrival rate function ofthe revealed queue and those of the concealed queue. We also show that when themarket size is sufficiently small (resp. large), the server always conceals(resp. reveals) the queue regardless of the realized service quality. In amedium-sized market, however, the server's optimal commitment strategy is oftenmixed, that is, randomized over queue disclosure and concealment. These resultsremain quite robust no matter whether customers are individual decision makersor act as one to maximize their total utility. We also find that, due to thechange of the optimal disclosure strategy, customers' total utility when theybehave collectively can be less than that when they behave individually. Whenthe server is a social planner, randomizing revealing and concealingthequeuemaystrictly dominates the strategy of completely revealing the queue on the wholerange of the market size.

主持人：王金亭教授，12BET副经理

报告时间：2020年12月15日（周二）上午10点

报告地点：腾讯会议ID：163 263 517，会议密码123456

主办单位：12BET