Machine Learning Approach to Uncover How Players

主讲人: Michihiro Kandori (Professor, University of Tokyo)

主持老师: （经院）吴泽南、石凡奇；（国发院）胡岠

参与老师: （经院）胡涛、吴泽南、石凡奇；（国发院）汪浩、胡岠

时间：2022年3月24日（周四）10:30-12:00

地点：ZOOM会议（会议号附在文末）

报告摘要：How do people learn to play a mixed strategy equilibrium? We analyze this issue in a large data set where more than 5,000 subjects played a game that has a highly non-trivial mixed strategy equilibrium. We employed some machine learning models to fit the data and found that some of them substantially improve upon a leading behavioral model of learning.  We also examine what kind of insights can be gained from the estimated machine learning models.

主讲人简介：

1989 Ph. D, Stanford University

1989 Assistant Professor, University of Pennsylvania

1990 Assistant Professor, Princeton University

1992 Associate Professor, University of Tokyo

1999 Professor, University of Tokyo

Michihiro Kandori is known for his pioneering contributions to the theory of repeated games and evolutionary game theory. He is Fellow of Econometric Society and currently Executive Vice President of Game Theory Society.

会议号：830 0424 9904

密码：607650