REQUIRED TEXTBOOK: course notes.

REFERENCE TEXTBOOKS: Algorithmic Game Theory, Nisan, Roughgarden, Tardos, and Vazirani, eds.

COURSE COORDINATOR: JASON HARTLINE (See Prof. Hartline's course page: http://www.eecs.northwestern.edu/~hartline/courses/algorithmic-mechanism-design/)

COURSE GOALS: The goal of this course is to familiarize graduate students (and advanced undergraduates) with the current state-of-the-art in algorithmic mechanism design. Through lectures and reading of classic research papers, students will familiarize (if necessary) themselves with foundational theory from both approximation algorithms and auction theory. Through lectures and the reading of recent research papers, students will study the paradigmatic problems in the new field of algorithmic mechanism design. Students will explore at least one topic area deeply through readings.

DETAILED COURSE TOPICS:

Nash equilibrium, Bayes-Nash equilibrium, dominant strategy equilibrium, games of incomplete information, single-item auctions, single-parameter mechanism design, welfare maximization and approximation (Bayesian and prior-free), revenue maximization and approximation (Bayesian and prior-free), revenue equivalence, all-pay mechanisms, structured procurement, multi-dimensional mechanism design, posted pricing, combinatorial auctions.

PREREQUISITES: Students should either have a solid algorithms background (E.g., EECS 336) or a solid auction theory background (a prior course in auction design or game theory).

ASSIGNMENTS: paper reading, problem solving, survey paper.

 

more news