6 ECTS credits
168 h study time

Offer 1 with catalog number 1024251ANR for all students in the 1st semester at a (A) Bachelor - preliminary level.

Semester
1st semester
Enrollment based on exam contract
Impossible
Grading method
Grading (scale from 0 to 20)
Can retake in second session
Yes
Enrollment Requirements
Students must have followed Machine Learning, Introduction to Artificial Intelligence, Mathematics: Calculus and Linear Algebra, Probability Theory and Statistics, Algorithms and Data Structures 1,Structure of Computer Programs 1, Logic and Formal Systems, Discrete Mathematics, AI Programming Project before they can enroll for Bayesian Methods
Taught in
Dutch
Faculty
Faculty of Sciences and Bioengineering Sciences
Department
Computer Science
Educational team
Pieter Libin (course titular)
Activities and contact hours
24 contact hours Lecture
24 contact hours Seminar, Exercises or Practicals
24 contact hours Independent or External Form of Study
Course Content

This course covers Bayesian methods: data-driven methods that apply direct statistical analysis to the data. These techniques use Bayes' theorem to combine prior knowledge with observations (i.e., data points) from reality.

To ensure a solid understanding of these techniques, we build up the foundational concepts carefully (i.e., prior, likelihood, integration in the denominator, posterior, Bayes’ theorem, and inference). Based on these foundations, we explore more advanced concepts such as conjugate priors and Jeffreys priors, and study methods for evaluating the quality of models.

From there, we move on to Markov Chain Monte Carlo (MCMC), which we approach formally, intuitively, practically, and theoretically (including proofs). We conclude with advanced Bayesian techniques, including Bayesian Networks, Bayesian hierarchical models, and Bayesian linear regression.

Additional info

Over the course of 12 weeks, both theoretical and practical aspects are taught (HOC) and practiced during exercise sessions (WPO). During these sessions, students will have the opportunity to apply theory and gain experience in programming Bayesian algorithms and performing Bayesian analyses.

Learning Outcomes

General competencies

  • In-depth knowledge and understanding of Bayesian statistics, inference and machine learning.
  • Be able to formulate, apply, implement and validate Bayesian models.
  • Be able to create a project plan to solve a typical learning problem with Bayesian techniques.
  • Be able to independently update the knowledge acquired and to tackle new problems in science and applications.

Grading

The final grade is composed based on the following categories:
Practical Exam determines 50% of the final mark.
Other Exam determines 50% of the final mark.

Within the Practical Exam category, the following assignments need to be completed:

  • exam practical with a relative weight of 1 which comprises 50% of the final mark.

    Note: Project work leading to a paper of maximum 5 pages

Within the Other Exam category, the following assignments need to be completed:

  • exam other with a relative weight of 1 which comprises 50% of the final mark.

    Note: Exam with exercises (written) and theory (oral)

Additional info regarding evaluation

- Written exam: 50% of the total score
- Project work: 50% of the total score

The project consists of an individual assignment in which students develop a design and programming task. Students will submit a Python notebook and present their work. During the presentation, the teaching staff will also ask questions to assess understanding of the methods and techniques covered.

To take the written exam, students must register for the project. To pass the course, the student must pass both components: if not, the final score will reflect the lowest of the two. If the student fails the written exam but passes the project, the project score can be carried over to the resit exam (second session).

The use of Generative AI (genAI) is strictly limited to enhancing the clarity and readability of your written work. It may be used for grammar correction, sentence restructuring, and improving the flow of text. However, genAI cannot be used to generate ideas, conceptualize project components, or write any part of the required code. All project work must reflect your own original thought process and problem-solving skills. Any unauthorized use of genAI beyond these limitations will be considered a violation of academic integrity. If genAI is used, this must be declared.

Allowed unsatisfactory mark
The supplementary Teaching and Examination Regulations of your faculty stipulate whether an allowed unsatisfactory mark for this programme unit is permitted.

Academic context

This offer is part of the following study plans:
Bachelor of Artificial Intelligence: Default track (only offered in Dutch)