MSV Incognito Knowledge Base

Intelligent systems introduces the student to computational agents, which act and interact flexibly and autonomously in order to achieve some goal. This course discusses the basic concepts and methods from agent technology. Topics that are covered include: characteristics of an agent, agent architectures, cooperation and competition among agents, and agent communication. This course contains a practical part where the students work in teams on a concrete application and implementation of an agent system. After completing this course the student will be able to judge whether it is beneficial to use agent technology over other approaches. The student will have a basic understanding of agent technology, which allows the student to apply agent technology to practical problems.

The course contents are (2022):

1. Intelligence vs Autonomy

2. Expert Systems

3. Agent Oriented Design

4. Genetic Algorithms

5. Multi Agent Coordination

6. Multi Agent Learning (x2)

7. Deep Learning for Agents

8. Reaching Human Level Intelligence (+ guest lecture?)

(2022) This course is currently taught by Kurt Driessens. There are two lectures on MARL, Multi Agent Reinforcement Learning (or topic 6 in the list above), make sure to be there when he goes through the lectures, as it is very difficult to study from just the slides and it's also difficult to find anything online about it. Do note that it wasn't that present on the exam in 2022. IS is a course where you have to study quite a bit, so make sure you start on time.

(2022) 80% of your grade is a closed book exam. The exam of IS is partly multiple choice and partly normal questions, around 50/50. If you mark a question wrong in the multiple choice part points get reduced. This makes the exam pretty tricky.

20% of your grade is determined by one out of three choises: 1. pick an intelligent system and do a 10 min video on it explaining it in detail (in pairs), 2. Create an intelligent system and make a video about it (3-4 students), 3. Write a report on 2 out of 4 labs (in pairs)

There are 4 labs which are mandatory but non-graded.

None.

• Mike Wooldridge (2009, 2nd edition): An Introduction to Multi Agent Systems, Michael Wooldridge, John Wiley & Sons Ltd, ISBN-13: 978- 0470519462
• Stuart Russell and Peter Norvig (2010, 3rd edition), Artificial Intelligence, A Modern Approach, Prentice-Hall, ISBN-13: 9780136042594.