600.473 Algorithmic Game Theory - Spring 2016

Instructor: Michael Dinitz
Lecture: T Th 3:00 - 4:20 pm, Maryland 109
Office Hours: By appointment

Syllabus

Course Description: This course provides an introduction to algorithmic game theory: the study of games from the perspective of algorithms and theoretical computer science. There will be a particular focus on games that arise naturally from economic interactions involving computer systems (such as economic interactions between large-scale networks, online advertising markets, etc.), but there will also be broad coverage of games and mechanisms of all sorts. Topics covered will include a) complexity of computing equilibria and algorithms for doing so, b) (in)efficiency of equilibria, and c) algorithmic mechanism design. Students may receive credit for 600.473 or 600.673, but not both. Pre-req: 600.363/463 or permission.

Textbook: Algorithmic Game Theory, Nisan, Roughgarden, Tardos, and Vazirani, Cambridge University Press, 2007. username: agt1user, password: camb2agt.

Schedule

The following is a tentative schedule. As the class proceeds, there will likely be changes and modifications.

Date Topic Reference Notes
 Jan 26   Snow day!       
 Jan 28   Introduction. What is AGT? Basic Game Theory   Chapter 1, Lecture notes from Stanford 
 Feb 2   Two-Player Zero-sum Games   Chapters 1.4.2, 3.1, 3.2
 Feb 4   Algorithms for Two-Player Games: Lemke-Howson   Chapter 3.3, 3.4, 3.6  HW1 released 
 Feb 9   Hardness of Computing Nash   Chapter 2, Lecture Notes from Stanford 
 Feb 11   Potential Games and PNE; Hierarchy of Equilibria   Chapters 1.3.4, 1.3.6, 19.3.1, 19.3.2, Lecture Notes from Stanford   HW1 due
 Feb 16   No-Regret Algorithms and Coarse Correlated Equilibria   Chapter 4, Lecture Notes from Stanford 
 Feb 18   No-Swap-Regret and Correlated Equilibria   Chapter 4, Lecture Notes from Stanford 
 Feb 23   Intro to Inefficiency of Equilibria   Chapter 17 
 Feb 25   Non-Atomic Routing Games   Chapter 18   HW2 released
 Mar 1   Atomic Routing Games   Chapter 18 
 Mar 3   Smooth Games, Facility Location   Notes from Stanford, Chapter 19.4 
 Mar 8   Location Game, Connection Game   Chapters 19.3, 19.4 
 Mar 10   Connection Game, Strong Nash   Chapter 19.3, Lecture Notes from Stanford  HW2 due 
 Mar 15   Spring Break   
 Mar 17   Spring Break   
 Mar 22   Load Balancing   Chapter 20 
 Mar 24   Intro to Mechanism Design   Chapters 9.3.1, 9.3.2, 9.3.5, Lecture Notes from Stanford   HW3 released 
 Mar 29   Myerson's Lemma for Single Parameter Environments   Chapters 9.4.1-9.4.2 and 9.5.4-9.5.6, Lecture Notes from Stanford 
 Mar 31   Knapsack Auctions, Revelation Principle   Chapters 9.4.3, 12.1, 12.2. Lecture Notes from Stanford 
 Apr 5   Revenue-Maximizing Auctions   Chapters 13.1, 13.2. Lecture Notes from Stanford 
 Apr 7   "Simple" Approximate Auctions   Chapters 4.2, 5.2, 5.2.1 of Hartline's book, and Lecture Notes from Stanford   HW3 due, HW4 released 
 Apr 12   Bulow-Klemperer, Multi-Parameter Environments and VCG   Chapter 5.2.1 of Hartline's book, Chapter 9.3.3 of AGT book, Lecture Notes from Stanford 
 Apr 14   Combinatorial Auctions, Spectrum Auctions   Chapters 11.1, 11.2. Lecture Notes on spectrum auctions from Stanford. 
 Apr 19   Mechanism Design Without Money   Chapters 10.3, 10.4 
 Apr 21   Online Auctions   Slides from Maryland   HW4 due
 Apr 26   Voting Schemes and Social Choice   Chapter 9.2  
 Apr 28   Project presentations   

Assignments

Please submit homeworks using Gradescope.

Announcements

Project Information

This class will have a final project which will consist of each student reading a modern research paper on algorithmic game theory and giving a 20-minute presentation to the class. Good conferences to look for interesting papers are EC (Electronic Commerce), WINE (Workshop in Internet and Network Economics), STOC, FOCS, and SODA.

Additional Resources

Some texts that are related to topics we will cover. They are useful for further study in this field. Similar courses with nice notes.