Syllabus for Reinforcement Learning

Förstärkningsinlärning

Syllabus

  • 5 credits
  • Course code: 1RT745
  • Education cycle: Second cycle
  • Main field(s) of study and in-depth level: Technology A1N, Embedded Systems A1N

    Explanation of codes

    The code indicates the education cycle and in-depth level of the course in relation to other courses within the same main field of study according to the requirements for general degrees:

    First cycle

    • G1N: has only upper-secondary level entry requirements
    • G1F: has less than 60 credits in first-cycle course/s as entry requirements
    • G1E: contains specially designed degree project for Higher Education Diploma
    • G2F: has at least 60 credits in first-cycle course/s as entry requirements
    • G2E: has at least 60 credits in first-cycle course/s as entry requirements, contains degree project for Bachelor of Arts/Bachelor of Science
    • GXX: in-depth level of the course cannot be classified

    Second cycle

    • A1N: has only first-cycle course/s as entry requirements
    • A1F: has second-cycle course/s as entry requirements
    • A1E: contains degree project for Master of Arts/Master of Science (60 credits)
    • A2E: contains degree project for Master of Arts/Master of Science (120 credits)
    • AXX: in-depth level of the course cannot be classified

  • Grading system: Fail (U), Pass (3), Pass with credit (4), Pass with distinction (5)
  • Established: 2020-02-25
  • Established by:
  • Revised: 2022-01-25
  • Revised by: The Faculty Board of Science and Technology
  • Applies from: Autumn 2022
  • Entry requirements:

    120 credits including Probability and Statistics, Linear Algebra II, Single Variable Calculus, Programming and Automatic Control I. Proficiency in English equivalent to the Swedish upper secondary course English 6.

  • Responsible department: Department of Information Technology

Learning outcomes

On completion of the course, the student shall be able to:

  • Explain possibilities and limitations of reinforcement learning.
  • Analyze relevant applications, decide if they can be formulated as a reinforcement learning problem, and in such case define it formally.
  • Implement, use and evaluate central algorithms for reinforcement learning.

Content

This course gives a solid introduction to the modern tools used to devise, implement and analyse reinforcement learning algorithms. The course covers Markov decision processes, planning by dynamic programming, model-free prediction and control, the trade-off between exploration and exploitation, function approximations and policy gradient methods. It also introduces deep reinforcement learning. Applications discussed in the course include classical control problems such as the inverted pendulum, as well as robotics and computer games.

Instruction

Lectures, seminars, computer labs.

Assessment

Oral and written examination of assignments (2 credits), and a written exam (3 credits).

If there are special reasons for doing so, an examiner may make an exception from the method of assessment indicated and allow a student to be assessed by another method. An example of special reasons might be a certificate regarding special pedagogical support from the disability coordinator of the university.

Other directives

The course cannot be included in the same degree as 1RT747 Reinforcement Learning.

Syllabus Revisions

Reading list

Reading list

Applies from: Autumn 2022

Some titles may be available electronically through the University library.

  • Sutton, Richard S.; Barto, Andrew G. Reinforcement learning : an introduction

    Second edition: Cambridge, MA: The MIT Press, [2018]

    Find in the library

Print syllabus and reading list
Print