News
Currently, no news are available
Welcome to the Advanced topics in causality and causal ML research seminar - Summer 26
In data science and real-world machine learning, there are many issues that are often neglected in standard machine learning courses. Many tasks are inherently trying to answer causal questions and gather actionable insights, even when there is not enough data to draw causal conclusions. Moreover, integrating causal reasoning in state-of-the-art AI can help improve the robustness and generalizability of current approaches, as well as imbue them with the strong theoretical guarantees typical of causality research.
In this seminar we will focus on the intersection of causality and machine learning research, by reading recent publications, discussing them and implementing extensions of current approaches in small groups. In particular, we will investigate cutting-edge research in three different research directions: 1) causal representation learning, 2) causal discovery and 3) downstream tasks, in particular causality-inspired ML/RL.
Disclaimer
Since this is the first time I teach in Germany (I had taught extensively at the University of Amsterdam, but the system is quite different), I might need to adapt/change things as we go. I'm also quite new (officially started at UdS on 1st March), so I might need to learn the UdS systems during the course and I might be less effective than I would like.
Tentative Timetable:
We will start with 4 lectures given by the lecturer (prof. dr. Sara Magliacane) on all three main topics: causal discovery, causal representation learning and downstream tasks/causality-inspired ML/RL. Please note that three of these lectures will be online (see the zoom link below).
After that we will divide in groups and each group will focus on reproducing a paper and proposing a small extension. We will be having 5 progress meetings with the groups on Thursdays at 10-12, with some weeks cancelled due to holidays or conferences (see which weeks in the timetable below).
Finally in the last week (July 13-17th 2026) each group will present their project and submit a Jupyter notebook.
| When | Where | Activity |
|
April 16th at 10-12 |
SR106 | Inperson lecture - Course logistics + Intro to causality |
| April 23rd at 10-12 | https://uva-live.zoom.us/j/63564721506 | Online lecture - Causal discovery |
| April 30th at 10-12 | https://uva-live.zoom.us/j/63564721506 | Online lecture - Causal representation learning |
| May 7th at 10-12 | https://uva-live.zoom.us/j/63564721506 | Online lecture - Causality-inspired ML/RL |
| May 21st at 10-12 | SR106 | Progress meeting with groups |
| May 28th at 10-12 | SR106 | Progress meeting with groups |
| June 11th at 10-12 | SR106 | Progress meeting with groups |
| June 25th at 10-12 | SR106 | Progress meeting with groups |
| July 2nd at 10-12 | SR106 | Progress meeting with groups |
| July 13-17th (TBD) | SR106 | Group Presentations |
Suggested reading per week
Week 1: Intro to causality: first two chapters of the Causality Primer by Pearl et al. 2021
Week 2: Causal discovery: Chapter 5 (and possibly 6) of Spirtes, Glymour, Scheines, Causation, Prediction, Search + survey by Glymour et al (2019)
Week 3: Causal representation learning: Towards Causal Representation Learning by Schölkopf et al. 2021
Week 4: Downstream tasks/causal ML: Causal Machine Learning survey by Kaddour et al. 2022 (includes also some of the previous topics)
Grading
The final evaluation will be based on a group project presentation and a Jupyter notebook for the project.
Presentation: Each group member is expected to present for 10 minutes and answer to questions for 5 minutes.
Jupyter notebook: The notebook should clearly mention which parts were done by which group member, and provide reproducible code for the project, as well as some theoretical explanation, similar in style (but not content) to this tutorial https://uvadlc-notebooks.readthedocs.io/en/latest/tutorial_notebooks/DL2/Causality_and_CRL/citris-tutorial.html
The grading will be 50% presentation and 50% Jupyter notebook.
Additional material
Related basic course (covers many aspects of causality beyond what we will do here): Causality course at the University of Amsterdam https://video.uva.nl/playlist/dedicated/0_ug2t9vyf/
Relevant basic textbooks (mostly give the basic background for each topic):
Pearl, Glymour, Jewell, Causal Inference in Statistics: A Primer