Course content

The course gives an overview of the main aspects of multi-agent systems, for example coordination of the behaviour of various agents sharing the same environment. Both cooperative and selfish agents and interactions between them will be discussed. Central to the course is interaction protocols such as auctions, negotiations, etc. Game theory will be a significant part of the course. A practical part of the course is assignments/projects involving implementation of various aspecs of multi-agent systems and game theory.

Learning outcome

Knowledge:
The candidate will get knowledge of :
- main principles of distributed AI
- which techniques from artificial intelligence (AI) can be used in distributed AI environments
- various agent types and their characteristics
- game theory concepts relevant to multiagent systems
- how do agents take strategical decisions
- agent communication languages and interactions between agents.

Skills:
The student can:
- decide which agent types can be used in different problems/tasks
- design agent interaction (e.g., negotiation) protocols
- design information/knowledge models and reasoning algorithms for agents.

General competence:
The student can:
- develop intelligent agents and build multiagent systems.

Learning methods and activities

Lectures, colloquia, self study, exercises. Portfolio assessment is the basis for the grade in the course. The portfolio includes a final written exam (70%) and exercises/projects (30%). The results for the parts are given in %-scores, while the entire portfolio is assigned a letter grade. If there is a re-sit examination, the examination form may be changed from written to oral.

Compulsory assignments

• mandatory assignments

Recommended previous knowledge

TDT4136 Introduction to Artificial Intelligence or equivalent.

Course materials

Textbook: Wooldridge, M.J.: An Introduction to Multiagent Systems. A set of papers: Will be announced at the start of the course.

Credit reductions