TFSS25 - Reinforcement Learning

Course Code:	TFSS25
Confirmed:	Nov 28, 2024
Valid From:	Aug 31, 2026

Education Cycle:	Second-cycle level
Disciplinary domain:	Technology
Subject group:	Computer Technology
Specialised in:	A1F Second cycle, has second-cycle course/s as entry requirements
Main field of study:	Computer Science

Intended Learning Outcomes (ILO)

On completion of the course the student shall:

Knowledge and understanding

display knowledge of central theories, algorithms, and performance criteria in Reinforcement Learning (RL)
display an understanding of possibilities and limitations in RL
display knowledge of common tools and frameworks used in RL
display an understanding of seminal contributions in the RL research literature

Skills and abilities

demonstrate skills in formulating problems as a Markov Decision Process (MDP)
demonstrate skills in choosing appropriate RL algorithms for specific MDPs
demonstrate skills in developing and implementing RL solutions to various problems using common tools and frameworks

Judgement and approach

demonstrate sound judgement in formulating problems as MDPs, selecting appropriate RL algorithms for solving specific MDPs, and evaluating solutions across multiple performance criteria.

Content

The quest to fully realize the potential of Artificial Intelligence (AI), requires autonomous systems that can learn to make good decisions by interacting with their environment. Reinforcement learning is a paradigm that meets these requirements, and can be applied to various tasks, including game-playing, healthcare, economics, and robotics. This course gives a solid introduction to reinforcement learning with its core approaches and challenges, and is structured around several lectures, assignments, and a project.

The course includes the following elements:

Markov Decision Processes (MDPs)
Model-based and model-free prediction and control
On-policy and off-policy methods
Monte Carlo, Temporal Difference, Policy-Gradient, and Actor-Critic methods
The exploration versus exploitation trade-off, including regret
The bias versus variance trade-off, including stability
Function approximation, including Deep Reinforcement Learning
Imitation Learning and Reinforcement Learning with multi-agent interactions

Type of instruction

Lectures, exercises, and seminars.

Language of instruction is English.

Entry requirements

Passed courses at least 90 credits within the major subject Computer Engineering, Computer Science, Electrical Engineering (with relevant courses in Computer Engineering) or equivalent, or passed courses at least 150 credits from the Computer Science and Engineering programme, and taken courses in Artificial Intelligence, 7,5 credits, Machine Learning, 7,5 credits and Deep Learning, 7,5 credits or equivalent. Proof of English proficiency is required.

Examination and grades

The course is graded 5, 4, 3 or U.

Registration of examination:

Name of the Test	Value	Grading
Assignment ¹	5 credits	5/4/3/U
Project	2.5 credits	G/U

¹Determines the final grade of the course, which is issued only when all course units have been passed.

Course literature

Please note that changes may be made to the reading list up until eight weeks before the start of the course.

Title: Reinforcement Learning, 2nd Edition
Author: Richard S. Sutton and Andrew G. Barto
Publisher: Bradford Books, 2018
ISBN: 9780262039246

Title: Grokking Deep Reinforcement Learning
Author: Miguel Morales
Publisher: Manning, 2020
ISBN: 9781617295454

COURSE SYLLABUS