6.853 Topics on Algorithmic Game Theory and Data Science

SHARE:

Graduate Level
Units: 3-0-9
Prerequistes: 6.006 or equivalent
Instructors:  Professor Constantinos Daskalakis (costis@csail.mit.edu) and Vasilis Syrgkanis (vasilis@cs.cornell.edu)
Schedule:  WF11:30-1, room 6-120
 
Subject Description:
 
The course will present topics at the interface of Game Theory, Economics, Algorithms and Learning. We will study this interface from three angles: (i) We will present the foundations of Game Theory and their intimate connection to duality theory and online learning; (ii) We will present the fundamentals of auctions and of mechanism design, studying the learnability of good auctions and mechanisms from data through the perspective of provably-approximately-correct (PAC) learning; and (iii) We will present the basics of Econometrics through the prism of density estimation. Thrust (i) will cover the basics of strategic behavior, equilibria, duality theory, online learning, and price of anarchy. Thrust (ii) will present the basics of mechanism design, revenue optimization, and PAC learning, turning to the study of simplicity, learnability, and approximation tradeoffs in mechanism design. Thrust (iii) will present the basics of Econometrics and Statistics, turning to inference in games and auctions. The running examples in this course will be motivated by applications in online advertising and online market design.