Individual Versus Social Optimization In Statistical Estimation

Abstract

A number of estimators participate in a non-cooperative game where the data they collect are reported to a central planner who takes decisions. She in her turn decides on the estimates to be imposed on the participants. Because they are aware of this, they may behave strategically and misreport. We exemplify this decision making process where the James-Stein’s estimator is used by the central planner. In particular, we consider the resulting optimization or Nash equilibria reporting strategies in a number of variations of this game.

Biography

Moshe Haviv got his B.Sc. in mathematics from Tel  Aviv University in 1979, and his M.A. and Ph.D in Operations Research from Yale University in 1982 and 1983, respectively. He joined the department of Statistics at the Hebrew University in 1984. He has been  a Professor since 2002 and until his retirement in 2020. He served as Head of Department from 2008 through 2012. In 2010 he joined the Center for the Study of Rationality.  He also  taught at the University of British Columbia and at the University of Sydney. Moshe was the president of the Operations Research Society of Israel from 2011 to 2014. Currently he is a Professor at the School of Data Science at the  Shenzhen campus of the Chinese University of Hong Kong. His research areas are queueing systems in general and strategic decision making in queues, in particular.  Other areas of interest are numerical issues in Markov chains and Markov decision processes. Among his publications, a book titled "To queue or not to queue: Equilibrium behaviour in queueing systems", co-authored with Refael Hassin (Kluwer, 2003). He also published a textbook "Queues: A Course in Queueing Theory" (Springer, 2013).  He has released a textbook titled "Linear Algebra for Data Science" (World Scientific, 2023). He also published more than 90 research papers. See  http://pluto.mscc.huji.ac.il/~haviv/.