CSL Master Calendar

Back to Listing

DCL Seminar: Ali Jadbabaie - An Axiomatic Foundation for Non-Bayesian Learning in Networks

Event Type
Seminar/Symposium
Sponsor
Decision and Control Laboratory, Coordinated Science Laboratory
Location
CSL Auditorium, Room B02
Date
Oct 28, 2015   3:00 pm  
Speaker
Ali Jadbabaie, Ph.D.
Contact
Linda Meccoli
E-Mail
lmeccoli@illinois.edu
Phone
217-333-9449
Views
22
Originating Calendar
CSL Decision and Control Group

Decision and Control Lecture Series

Coordinated Science Laboratory

 

“An Axiomatic Foundation for Non-Bayesian Learning in Networks”

Ali Jadbabaie, Ph.D.
Visiting Professor
Interim Director of the
Sociotechnical Systems Research Center
MIT
 

Wednesday, October 28, 2015

3:00 p.m. to 4:00 p.m.

CSL Auditorium (B02)

____________________________________________________________________________________________________________________________________________________________________

 

 

“An Axiomatic Foundation for Non-Bayesian Learning in Networks” 

 

Abstract:

Rational learning postulates that individuals incorporate new information into their beliefs in a Bayesian fashion. Despite its theoretical appeal, this Bayesian learning framework has been criticized on the basis of placing unrealistic computational demands on the agents. Furthermore, experiments have shown that the way agents update their beliefs in networked settings is often inconsistent with predictions of Bayesian learning models.  Motivated by these issues, A large body of literature has emerged that proposes a series of non-Bayesian updates that are often inspired by the linear (consensus) learning model of DeGroot. However, a systematic framework that captures behavioral deviations of such updates from Bayesian learning has been lacking.

In this talk, I will present a theoretical foundations for non-Bayesian aggregation of beliefs by taking an axiomatic approach that introduces a set of axioms about the agents’ behaviors. Rather than proposing functional form for the belief updates, I will present behavioral axioms from which updates are derived. The first behavioral assumption is a notion of imperfect recall, according to which agents take the current belief of their neighbors as sufficient statistics, ignoring how and why their current opinions were formed. Next, I will impose a set of behavioral axioms that are all satisfied by Bayesian agents. I will show that a combination of these behavioral assumptions leads to an update that combines log-likelihood  ratios of the neighboring beliefs in a linear fashion.  Using the notion of group polarization from social psychology literature, I will provide conditions on these weights that lead to learning and mis-learning. I will  then discuss implications of relaxing some of the axioms not he functional form of the updates. If time permits, I will  also characterize the rate of convergence for this family of updates and describe how the rate depends on the network structure and information endowment of agents.

Joint work with Pooya Molavi (MIT Economics), Alireza Tahbaz Salehi (Columbia Graduate School of Business) , Shahin Shahrampour (Harvard), and Amin Rahimian (Penn ESE)
 

Bio:

Ali Jadbabaie is currently on leave from University of Pennsylvania at MIT as a visiting Professor, where he is also the Interim Director of the Sociotechnical Systems Research Center (SSRC) and the Associate Director of the newly formed MIT Institute for Data, Systems and Society (IDSS).  He received his B.S. with High Honors from Sharif University of Technology in Tehran, Iran, his M.S. in Electrical and Computer Engineering from the University of New Mexico in Albuquerque, and his Ph.D. in Control and Dynamical Systems from the California Institute of Technology (Caltech). He was a postdoctoral scholar at Yale University before joining the faculty of the University of Pennsylvania (Penn) in July 2002, where he is currently the Alfred Fitler Moore Professor of Network Science in the department of Electrical and Systems Engineering. He holds secondary appointments in the departments of Computer and Information Science as well as Operations, Information, and Decisions in the Wharton School. A faculty member in Penn’s General Robotics, Automation, Sensing & Perception (GRASP) Lab, is also the Co-founder and director of the Raj and Neera Singh Program in Networked & Social Systems Engineering (NETS) at Penn Engineering,  a new undergraduate interdisciplinary degree program focused on network science and engineering, operations research, social phenomena and social networks. He is also a faculty member of The Warren Center for Network & Data Sciences at Penn and a faculty affiliate of the Center for Technology, Innovation and Competition at Penn Law.  He is the inaugural editor-in-chief of IEEE Transactions on Network Science and Engineering, a new interdisciplinary journal sponsored by several IEEE Societies. He is a recipient of an NSF Career Award, an ONR Young Investigator Award, the O. Hugo Schuck Best Paper Award from the American Automatic Control Council, and the George S. Axelby Best Paper Award from the IEEE Control Systems Society. He is also an IEEE Fellow.  His current research interests include decision and control theory with a focus on distributed optimization and control, collective behavior, network science, and the study of social and economic networks.

link for robots only