Bayesian Statistics and Data Analysis, Herbstsemester 2026, ETH Zürich
 |
 |
This course is taught in the fall term 2026 at ETH Zurich By Patrick Meyers, assisted by Uddipta Bhardwaj and Giada Badaracco. The material covered in the course is listed in the syllabus and laid out in detail in COURSE_CONTENT_AND_RESOURCES.md; it is influenced by previous generations of this course. It ranges from an introduction to probability theory up to Hamiltonian Monte Carlo and simulation-based inference. Since the course aims to focus mostly on showing how to use statistical analysis tools, the various topics are introduced only relatively briefly. There are many great books and courses on this topic, which go into more detail and which I encourage you to look at before the lectures. The resources listed here are all available online, either directly or through the ETH library. For example:
- Data Analysis: A Bayesian Tutorial, 2006. The textbook that we will follow the closest throughout this course.
- Bayesian Data Analysis, Gelman, 2013 — ETH library, link. The title says it all.
- Information Theory, Inference, and Learning Algorithms, MacKay, 2003 — link. Heavy on the information theory but also covers inference methods nicely. The exercises come with solutions.
COURSE_CONTENT_AND_RESOURCES.md is the map of the course: the week-by-week outline of all twelve topics with the exercises attached to each, plus the literature and links to the other courses we draw on. If you only read one file in this repository, read that one.
| Where | What |
|---|---|
| Start with setting up SSH and forking the repo | If you're new to the course, this is the first step and we cover this in the first lecture. |
| COURSE_CONTENT_AND_RESOURCES.md | Course outline, week by week, and the reading list. |
| SYLLABUS_HS2026.pdf | The official syllabus for the semester. |
| lectures/ | The lecture notebooks, which are the source for the slides and PDFs — currently week1_intro.ipynb. |
| slides/ | Rendered slides for the lectures — currently week1_intro.slides.html. |
| exercises/ | Exercise notebooks — currently week1_exercise_estimate_pi.ipynb. |
| exercise_solutions/ | Solutions to selected exercises — currently week1_pi_exercises_solution.ipynb. |
| course_tools/ | Plotting helpers, matplotlib styles, and the scripts that build the slides and PDFs. |
| environment.yaml | The conda environment for running everything here. |
| help | Infographics and other helpful accessories for git, python and more thanks to Michael Coughlin's repository. |
These will be populated as the course progresses through the term; at the moment only week 1 is in place.
This course stands on material generously shared by others:
- Tilman Tröster and Veronika Oehl, ETH Zürich — earlier generations of this course, which we follow closely (repository).
- Michael Coughlin, University of Minnesota — the Big Data in Astrophysics course, from which we borrow the applied, large-dataset perspective as well as the git cheat sheets and fork-syncing guides in our resource list (repository).
- Ben Farr, University of Oregon — computational physics, a source of project topics and datasets (repository).
See COURSE_CONTENT_AND_RESOURCES.md for the full set of courses and resources we draw on.