[Course] NOWE, Ann
Full text not available from this archive.
In this course the basics of Multi-agent systems is introduced, and links to other disciplines are discussed. Students should be able to apply these principles to differente applications. The main focus is on learning in Multiagent systems. The challenge that is studied is the one of an agent learning to behave in an environment where other agents are also learning at the same time. Multiagent learning is the result of a synergy of ideas from Game Theory and Reinforcement Learning. De basic concepts of both worlds are introduced, an overview of different techniques that join concepts into algorithms are discussed. The students should be able to apply these techniques to given case studies.
The course covers topics including: normal-form games, best-responses, Nash equilibria and its refinements, regret minimization, Markov decision processes, temporal difference learning, stochastic games, algorithms for learning equilibria, algorithms for learning best-responses, and algorithms for learning to coordinate.
BA computer science
BA mathematics option computer science
Level of Course: MSc
Status of Course: Running
Average Number of Students: 10-20
Length of Course: 1 semester
Lecturers: NOWE, Ann
Archive Staff Only: edit this record