CS-E5710: Bayesian Data Analysis
| Course name | Bayesian Data Analysis |
|---|---|
| Course code | CS-E5710 |
| Abbreviation | “BDA, Bayes” |
| Period | I-II |
| Lecturer | Aki Vehtari |
Description
After completing the course, students can explain core Bayesian concepts, outline the modeling process, formulate and recognize common statistical models, and implement/compare simulation methods (analytic, MCMC, etc.) to learn parameters, estimate model fit, and perform model comparison.
Course material
Official material
Book “Bayesian Data Analysis, 3rd ed”, lectures, videos, chapter notes, demos, assignment instructions, and the course website.
Extra material
Contents and workload
Overall workload
Weekly contents
The 13-week curriculum covers topics from basic Bayesian inference and Monte Carlo methods to Stan/HMC, hierarchical models, cross-validation, model comparison, decision analysis, and project evaluations.
Practicalities
Assignments (40%), exam studio e-exam (10%), and a project work with presentation (50%). A minimum of 50% of points is required from each component. Bonus points are available for active chat participation and answering time-usage questionnaires.
Related courses
Official prerequisites
Differential and integral calculus, basics of probability/statistics, and basic programming (R or Python). Matrix algebra is recommended.
Additional prerequisites
A course covering frequentist statistics (e.g., Statistical Inference) is useful for comparison.
More like this
Probabilistic Machine Learning, Artificial Intelligence.