C S
670
Multi-Agent Systems
Hours
3.0 Credit, 3 Lecture, 0 Lab
Introduction to fundamental concepts emphasizing current literature. Topics include game theory, repeated play games, Arrow's impossibility theorem, negotiation, search, and learning.
Understand key solution concepts in multi-agent systems
- Game theory formalism
- Nash equilibria
- Minimax solutions
- Pareto dominant solutions
- Strategically dominant solutions
Be able to model real problems using multi-agent formalisms
- Consequences, states, actions, goals, and utilities
- Extensive form games and normal form games
- Implementations of solution concepts: Nash equilibrium and minimax solutions
- Repeated-play games
- Evolutionary games
- Bio-inspired agent collectives
- Multi-agent learning
Analyze, implement, and communicate solutions
- Nash's theorem
- The folk theorem
- Replicator dynamics
- Repeated play equilibria
- Arrow's impossibility theorem
- Multi-agent learning
- Phases of nonlinear, bio-inspired collectives
Be literate in other solution concepts from multi-agent syst
- Market-based mechanisms: mechanisms and auctions
- Multi-agent search: Asynchronous dynamic programming and moving target search
- Reading current papers from the literature
Communicate solution and solution quality
- Write reports that explain theory, present experiments, and analyze results
- Recognize well-communicated ideas