讲座题目：Fractional Top Trading Cycle

　　报告人：张军

　　报告时间：2019年5月9日(周四)下午14点00分

　　报告地点：经管院B249

　　主办单位：数理经济与数理金融系

　　主持人：韩立宁

　　摘要: We generalize the Top Trading Cycle mechanism to solve random assignment problems. Specifically, we study the fractional endowment exchange problem in which each agent may own fractional amounts of multiple objects and each object may be owned by multiple agents. We propose a class of mechanisms based on a linear programming method. At every step, our mechanisms let agents point to most preferred objects and objects point to all of their owners. We use a linear equation system to describe how to trade the network generated at every step. Interestingly, the equation system is an instance of the classical Leontief input-output model so that its solutions must exist. We provide an intuitive explanation of our mechanisms: at every step, there exist disjoint absorbing sets in the generated network and agents in each absorbing set trade endowments only among themselves. All of our mechanisms are individually rational and sd-efficient. We characterize those mechanisms satisfying desirable fairness properties including equal-endowment no envy and stronger notions. We apply the mechanisms to solve real-life problems including school choice with weak priorities and time bank.

　　简介：张军博士2017年获加州理工学院经济学博士学位，2008至2012年间就读上海财经大学经济学硕博连读班，2008年毕业于南京理工大学应用数学系。张军博士主要研究领域为博弈论、市场设计和匹配理论，在Journal of Economic Theory, Economics Letters上发表文章，担任Econometrica, Journal of Economic Theory, Economic Theory 等经济学国际期刊的匿名审稿人。