3 credits
Fall 2026 Lecture Upper Division
The field of algorithmic game theory (also known as economics and computation) has been expanding at a rapid pace in recent years, with computational reasoning playing a crucial role in transforming classical economics. An increasing number of real-world systems require careful design of algorithms where the inputs are provided by agents with different preferences. Applications include Google's ad auctions, spectrum auctions, Bitcoin, matching markets, and allocation of vaccines. This has inspired new models and algorithms and led to deep connections with machine learning through multi-agent learning and computational complexity. This course will introduce the area of Algorithmic Game Theory by studying games, solution concepts, complexity of equilibria, mechanism design, auction theory, fair division, social choice, learning dynamics, and markets. It will include both classical results and recent developments.
Learning Outcomes1Analyze and solve games (zero-sum, general-sum, potential games), identify stable solutions, and design algorithms for computing them where applicable.
2Design and analyze mechanisms that work well when the participants are strategic, for various problems including selling goods (auctions).
3Design and analyze algorithms with provable fairness guarantees.
4Explain concepts of social choice used in specific domains, e.g., elections.
5Analyze markets and design algorithms that market participants can use to achieve desirable outcomes in both static and repeated settings.
6Analyze multi-agent learning settings, where agents learn together in some environment but have different incentives, which influence their interactions.
Prerequisites
Restrictions
made by Purdue students.