Swedish web site
Clustering2026VT, 6.0 credits
|
|
Course plan
No of lectures
ca 6 + 1 workshop with guest lecturers
Recommended for
PhD students who have an applied or engineering background, e.g., machine
learning, signal processing, statistics, computer vision, and control.
The course may be particularly useful for student desiring to apply clustering
methods in their research for exploratory purposes or data analysis.
Pre-requisites
• Python programming
• Linear algebra
• Knowledge of probabilities/statistics
Organization
3 days of lecturing + one workshop with guest seminars. Examination will of projects will be done online.
Content
This course serves as an introduction to the clustering task. We will mainly
focus on the different approaches with which clusters can be modelled,
entailing generative and discriminative models, and detail how model selection
can be done within each realm. The scope will purposefully remain generic for
any type of data, and fly over a variety of clustering models with an emphasis
on the seminal models: KMeans and Gaussian Mixture Models.
Intended learning outcomes comprise:
• A global understanding of the different types of clustering models
• The capability to choose an adequate model selection strategy in
cluster analysis
• How to integrate constraints in clustering such as labels or variable
selection
• An overview of the most common clustering models and their limitations
• How clustering can be adapted in deep learning contexts
The course will be completed by expert guest lecturers having some practical
applications of clustering.
Literature
• Bouveyron, C., Celeux, G., Murphy, T. B., & Raftery, A. E. (2019).
Model-based clustering and classification for data science: with applications
in R (Vol. 50). Cambridge University Press.
(https://math.univ-cotedazur.fr/~cbouveyr/MBCbook/)
• Ohl, L., Mattei, P. A., & Precioso, F. (2025). A tutorial on
discriminative clustering and mutual information. ACM Computing Surveys.
Lectures
• Generative modelling
• Discriminative modelling
• Model selection and evaluation
• Constraints in clustering
• Beyond vanilla clustering, variants of the clustering task
Examination
The course will contain 3 different exercises.
• A peer-reviewed project
• A critical analysis of a paper on clustering with presentation
• A paper examination
Examiners
• Louis Ohl
• Sebastian Mair
Credits
6
Comments
The course will be announced on the Grapes network.
Three guests lecturers are currently under planning for the workshop.
Page responsible: Anne Moe
