Applications of Learning Theory in Algorithmic Game Theory
published: Aug. 20, 2015, recorded: July 2015, views: 4195
Report a problem or upload filesIf you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.
Algorithmic game theory is a field that uses and extends tools from economics and game theory to reason about fundamental computer science problems. The field is important both for its applications, which span the gamut from network routing to online advertising, and for its remarkably diverse and rich connections to other areas of theoretical computer science, including complexity theory and approximation algorithms. In this talk, we survey two ways in which definitions and tools from learning theory have been crucial to recent advances in algorithmic game theory. First, we outline a theory of robust bounds on the "price of anarchy" --- meaning approximation guarantees for game-theoretic equilibria --- that apply to all outcome sequences generated by no-regret learners playing a multi-player game. Second, we explain how to use concepts from learning theory to make traditional (Bayesian) optimal auction theory operational, replacing the practically problematic "common prior" assumption with a data-driven approach.
Link this pageWould you like to put a link to this lecture on your homepage?
Go ahead! Copy the HTML snippet !