CPSC455/555 // ECON425/563: Economics and Computation
Time: TTh, 2:30 - 3:45
Location: AKW 200
Assistant: Judi Paige (AKW 507A, Judi.Paige@yale.edu, 203-436-1267)
Instructor Office Hours: Thurs 11:30 a.m. to 12:30 p.m. in AKW 512, or
TA: David Costanzo
TA Office Hours: Weds 2:30-3:30pm in AKW 301, or by appointment
: Algorithmic Game Theory
, by Nisan, Roughgarden, Tardos,
& Vazirani - available at the Yale bookstore
Errata for the textbook can be found here
Note: Do not send email to Professor Feigenbaum, who suffers from RSI. Contact
her through Ms. Paige or the TA.
A mathematically rigorous investigation of the interplay of economic theory and
computer science with an emphasis on the relationship of incentive
compatibility and algorithmic efficiency. Particular attention is paid to the
formulation and solution of mechanism-design problems that are relevant to data
networking and Internet-based commerce. Suitable for mathematically inclined
advanced undergraduates and first- or second-year graduate students in Computer
Science, Economics, or closely related fields.
This course was last taught in Fall 2008; more information can be found here.
For additional background reading about algorithms, please see Kleinberg and
Tardos's book entitled "Algorithm Design." It is on reserve at the Becton
library. Chapters 2 through 6 are the most important. Slides for this book
are available here.
For additional background reading about interdomain routing, please see the
first three chapters of Vijay Ramachandran's PhD thesis.
HW1 (due in class on Tues, Sept 20):
Chapter 14: 2, 3, 6, 7
Chapter 19: 3, 4, 6, 9
HW2 (due in class on Tues, Oct 11):
Chapter 11: 1, 3, 5
Chapter 28: 2, 4, 5
HW3 (due in class on Thurs, Oct 27):
Chapter 23: 1, 2
Chapter 27: 1, 2, 3, 5
(due in class on Thurs, Nov 17):
Chapter 27: 4, 6, 7, 8
Chapter 26: 4, Z*
* 26.Z is not from the
textbook, but is related to the material of Chapter 26. Here's the problem:
- (a) Show how to use the simple model of Section 26.5 to predict the results
of a referendum in which 60% of the voters must vote "yes" for the ballot
proposition to pass. That is, give a Boolean function f that captures the
semantics of such a referendum and prove that it satisfies the sufficient
condition under which the equilibrium price of F will reveal the true value
- (b) For the Boolean function f that you provided in part (a), show how the
price-formation process reveals the value of f(x) on the input vector (0, 1,
0, 0, 1) and the uniform common prior. That is, give the bids of each agent
in each round and the clearing price in each round, give the equilibrium
price, and explain why the process terminates when it does.
- due in class on Tues, Nov 29
Exam 1 in CPSC 455/555 // ECON 425/563 will be given in class on Thursday,
October 13, 2011.
- It will be a closed book exam and will cover the following material:
- Notes and slides from classes on Sept. 1 through Oct. 6
- Homeworks 1 and 2
- The following parts of Algorithmic Mechanism Design, by Nisan et al:
- Sections 1.1, 1.2, and 1.3
- Sections 9.3 and 9.4
- Chapter 11 (except 11.3 and 11.7)
- Chapter 14
- Sections 19.1, 19.2, and 19.3.1
- Chapter 28
- Most of the exam problems will be similar (but not identical) to homework
problems. You may also be asked to provide definitions, examples, theorem
statements, and/or (short) proofs that are given in the textbook sections above.
Exam 2 in CPSC 455/555 // ECON 425/563 will be given in class on Thursday,
December 1, 2011.
It will be a 1.25-hour, closed-book exam and will cover the following material:
Notes and slides from classes on Oct. 11 through Nov. 17, inclusive (except for
the part of Oct. 11 that was devoted to review for Exam 1 and the guest lecture
on Nov. 8 by Georgios Zervas)
- Homeworks 3, 4, and 5
- The following parts of Algorithmic Mechanism Design, by Nisan et al:
- Sections 23.1, 23.2, 23.3
- Sections 26.4, 26.5
- Chapter 27
- The paper "Computation in a Distributed Information Market," by Feigenbaum,
Fortnow, Pennock, and Sami
- Sections 2.4, 2.5, 3.2, 3.3, and 3.4 of the lecture notes from Professor Parkes's cs186 at Harvard
- As on Exam 1, some of the exam problems will be similar (but not identical) to homework problems. You may also be asked to provide definitions, examples, theorem statements, and/or (short) proofs that are given in the lectures, notes, paper, and textbook sections and chapter listed above.
Tues, Sept 20: HW1 due
Tues, Oct 11: HW2 due
Thurs, Oct 13: Exam 1
[Fri, Oct 21: Drop date]
Thurs, Oct 27: HW3 due
Thurs, Nov 17: HW4 due
[Nov 19 - Nov 27: Thanksgiving break]
Tues, Nov 29: HW5 due
Thurs, Dec 1: Exam 2
Note: There is no final exam during exam period at the end of the semester.
Homeworks: 50% (5 homeworks, 10% each)
Exam 1: 25%
Exam 2: 25%
- September 1, 2011: Course Overview, Introduction. [slides]
- September 6, 2011: Introduction to game-theoretic concepts and their
relevance to routing and network formation. Nash equilibrium,
dominant-strategy equilibrium, price of anarchy, price of stability. See
Chapters 1 and 17 of Nisan et al.
- September 8, 2011: Price of Stability and Price of Anarchy in local and
global connection games. See Sections 19.2 and 19.3 of Nisan et al.
- September 13, 2011: Algorithmic mechanism design, strategyproofness,
second-price auctions, and lowest-cost-path mechanisms
- September 15, 2011: Distributed algorithmic mechanism design, interdomain
routing, and ex-post Nash equilibrium. See Chapter 14 of Nisan et al.
Thanks to Yale
Ramachandran, we have this introduction to
interdomain routing and the Gao-Rexford framework.
- September 20, 2011: VCG mechanisms, the Clarke pivot rule, combinatorial
auctions (CAs), and a strategyproof, approximately optimal, polynomial-time
mechanism for single-minded bidder CAs. See Sections 9.3, 11.1 and 11.2 of
Nisan et al.
- September 22, 2011: Guest lecture by Professor Dirk Bergemann of the Yale
Economics Department, based on the following papers:
- September 27, 2011: Bidding languages and communication complexity in CAs.
See Sections 11.4 and 11.6 of Nisan et al.
- September 29, 2011: Guest lecture [.pptx, .pdf] by Shaili Jain of the Yale
Computer Science department on iterative CAs and the query model. See
Section 11.5 of Nisan et al.
- October 4, 2011: More detailed look at material needed for HW2 problems
from Chapter 11: algorithmic techniques, demand queries, value queries, and
- October 6, 2011: More detailed look at material needed for HW2 problems
from Chapter 28: Generalized Second Prize (GSP) ad auctions, full-information
Nash Equilibrium conditions, and locally envy-free equilibrium conditions.
We also reviewed the relationship of the generalized English auction and the
GSP and took an interesting detour into a discussion of user privacy in
advertiser-supported search and email.
- October 11, 2011: Review for Exam 1. Introduction to Peer-To-Peer (P2P)
systems. See Sections 23.1 and 23.2 of Nisan et al.
- October 13, 2011: [Exam 1][Solutions]
- October 18, 2011: A minimal P2P model and introduction to the BitTorrent system. See
Section 23.3 of Nisan et al. and Section 3.2
- October 20, 2011: Extensive form games, subgame-perfect equilibria, and
repeated PD. See Section 27.2 of Nisan et al. and Sections 2.4 and 2.5
- October 25, 2011: Strategic analysis of BitTorrent. See Sections 3.2.3,
3.3, and 3.4
- October 27, 2011: Introduction to the Pay-Your-Dues (PYD) strategy for
repeated Prisoners' Dilemma and to peer prediction in reputation systems.
See Sections 27.3 and 27.4 of Nisan et al.
- November 1, 2011: More about peer prediction and reputation functions.
See Sections 27.4 and 27.5 of Nisan et al.
- November 3, 2011: Value sybilproofness and rank sybilproofness of
reputation functions; see Section 27.5 of Nisan et al. The optimality of
the PYD strategy in repeated Prisoners' Dilemma games with whitewashing; see
this paper by Friedman and Resnick.
- November 8, 2011: Guest lecture by Georgios Zervas of the Yale Computer
Science Department, based on this
- November 10, 2011: Introduction to prediction markets. See Section 26.5 of
al. and this
- November 15, 2011: Weighted-threshold functions and the Shapley-Shubik
market game (Theorems 5 and 6 of
- November 17, 2011: The complexity of the matching problem in Combinatorial
Prediction Markets; see Section 26.3 of Nisan et al. In the Shapley-Shubik
market game, n rounds of trading suffice for weighted threshold functions of
n Boolean variables; see Theorem 8
- November 29, 2011: Review for Exam 2
- December 1, 2011: [Exam 2][Solutions]