“數理時(shí)空”學(xué)術(shù)論壇第十三講
講座題目:具有止步的優(yōu)先權排隊:博弈理論分析及二維均衡策略(Priority Queues withBalking: Game-theoretic Analysis and Two-dimensional Equilibrium Strategies)
時(shí)間:2019年5月22日(周三)10:00
地點(diǎn):西校區理學(xué)樓404會(huì )議室
報告人:王金亭
報告人簡(jiǎn)介:
王金亭,北京交通大學(xué)教授,博士生導師,教育部新世紀優(yōu)秀人才。2000年畢業(yè)于中國科學(xué)院數學(xué)與系統科學(xué)研究院應用數學(xué)所,獲“運籌學(xué)與控制論”專(zhuān)業(yè)博士學(xué)位。研究領(lǐng)域為隨機運籌(隨機服務(wù)與運作管理、可靠性理論、供應鏈優(yōu)化與管理、排隊博弈論等)、應用概率統計。現任中國運籌學(xué)會(huì )可靠性分會(huì )理事長(cháng)、中國運籌學(xué)會(huì )隨機服務(wù)與運作管理分會(huì )副理事長(cháng)、中國運籌學(xué)會(huì )理事、北京運籌學(xué)會(huì )副理事長(cháng)等學(xué)術(shù)職務(wù)。2006年獲得教育部霍英東教育基金會(huì )第十屆高等院校青年教師獎,2011年入選教育部新世紀優(yōu)秀人才支持計劃,2018年獲得詹天佑鐵道科學(xué)技術(shù)獎專(zhuān)項獎。發(fā)表學(xué)術(shù)論文100余篇,包括80余篇SCI檢索論文。SCI引用500余次,SCI嚴格他引400余次,Google Scholar引用計1600余次,出版專(zhuān)著(zhù)1部(《排隊博弈論基礎》,科學(xué)出版社,獨著(zhù)2016)
講座內容:
In this talk, we consider an M/M/1 queueing system with apay-for-priority option, and study customers' joint decisions between joining/balkingand pay-for-priority. The equilibrium strategies are thus two-dimensional.First, we fully characterize the equilibrium structure and identify thePareto-dominant strategies of such a game analytically, under both theobservable and unobservable settings. Interestingly, the equilibrium structure,the system throughput, and the service provider's optimal price for prioritypremium can all be non-monotone in the service reward, which departs from theexisting models of priority queues without balking. In particular, we find thatan increase in service reward can actually hurt the firm's revenue (everythingelse being equal). Second, we compare the server's revenue between theobservable and the unobservable settings. We find that the service provider isbetter off with the observable setting when the system load is either low orhigh, but benefits more from the unobservable setting when the system load ismedium. The fact that the optimal setting switches twice as the system loadincreases, is rather interesting; we explain the intuitions behind it in thispaper. Finally, we demonstrate the implications of these findings by applyingour model framework to Papa John's Pizza, based on publicly availableinformation. Our analysis suggests that Papa John's could benefit fromproviding customers with wait information while slightly decreasing its fee forPapa Priority.
歡迎廣大師生參加!
理學(xué)院
2019年5月20日