Learning dynamics

The course addresses the question how agents learn to act in environment where there are other agents also learning how to act.  Hence it is a course on multi-agent learning. 

We address this question from the perspective of the learning or adaptation algorithm, which can be 

1. the perspective of an individual agent, where learning how to act occurs internally,
2. the perspective of the collective, where one examines how the  distribution of agent behaviours changes over time.

Essentially we focus in the first perspective on the basis and state-of-the-art of multi-agent reinforcement learning and in the second perspective on evolutionary dynamics. 

• In the first 5-6 weeks of this course, the basis of game theory, evolutionary dynamics and multi-agent RL are explained
• In the second part we dive into a number of more advanced topics to expand the knowledge in these topics and to prepare you further for the exam work.

At the end of each theory session, if feasible, you will have the chance to play with some simple coding examples.  These  sessions will also introduce assignments, which you need to complete at home and that will be graded individually.  The exam is the reproduction of a paper on the topic, for which you will provide a report, code and a presentation in the exam session. 

 

The aim of the course is to introduce the students to the field of learning in individual agents and learning in populations of agents and to prepare them for a Master thesis in these research areas.

He or she will learn the basic principles of both domains, the mathematical and computational methods and the typical problems they are trying to solve.

The students will also obtain a basic understanding of (evolutionary) game theory which will allow them to understand the standard literature in that field and the relevance of this domain to learning in general.

The students will obtain the skills to address independently problems within these fields. In addition, they will be capable of presenting their work to an audience of specialists and non-specialists.

Cours ayant celui-ci comme co-requis

All information for this course is provided on UV.

References, bibliography, and recommended reading

All information for this course is provided on UV.

Course notes

• Université virtuelle

Additional information

This course is taught jointly with the VUB

Contacts

Email: Tom.Lenaerts@ulb.be; Office Campus La Plaine, NO8-117 
Email: Ann.Nowé@vub.be ; VUB, Pleinlaan 9, 3rd floor

Campus

Outside campus ULB

Method(s) of evaluation

• Other

Other

1. During the course a number of homeworks will have to be made which will be evaluated individually. There are two deadlines for the assignments. 
2. The exam is performed in groups of 3 students and is the reproduction of a scientific work related to the topics of this course.  A number of papers will be proposed from which a selection can be made. For the exam you will provide a report, discussing the reproduction of the work, the code to produce the results and a presentation on the exam date. 

Mark calculation method (including weighting of intermediary marks)

50% of the total score is given for the assignments and 50% is counted for the group project. 

Language(s) of evaluation

• english