(Ph.D. in Computer Science, University of Milan, 2016)

Lecturer: Dr. Nicola Basilico

This course provides an introduction to multiagent systems by concentrating on modeling agents interactions by means of competitive games. The main objectives of this course are: conveying basic notions of game theoretical models, discussing in detail some of the algorithms for their resolution, and presenting some recent real-world applications.

• The calendar has been updated: the class of April 22th is postponed to May 10th — Nicola Basilico 2016/04/19 13:33
• Anybody interested in attending this course should send an email to nicola.basilico@unimi.it — Nicola Basilico 2016/04/12 09:52

Class sessions will be held from 9:00 to 12:15 (with breaks)

1. Introduction to Algorithmic Game Theory, self-interested agents, von Neumann-Morgenstern preferences and utilities, strategic form games definition and examples, players strategies (April 19th 2016)

The required assignment promotes research thinking within the scope of the addressed topics. The course will provide basic notions that students are encouraged to apply to any problem (preferably but not necessarily in their area of expertise) they find interesting or significant, the objective being that of identifying (and, optionally, pursuing) possible research challenges. These are the main (not mutually exclusive) lines of work:

• find a problem you are interested in and try to model it as game played by multiple agents; describe your formalization and discuss some benefits and research challenges that it would introduce;
• select one the topics introduced in the course and briefly review the recent state-of-art on it or on some of its applications envisioning possible future directions of development;
• develop one or more tools that can be useful for the understanding of the course topics or that can support some insights discussed in the previous lines.

The minimal deliverable for the course assignment is a short report of at least 3 pages (lncs LaTeX format). Cooperation and discussion between students is encouraged (especially if from different research backgrounds). Assignments can be done by parties of at most 3 students, the workload should be equally divided and the contribution of each one should be explicitly pointed out in the report.

Students that deliver the assignment will receive a grade on a scale from A to E. To successfully register the course, a grade of at least C is required.