November 30th - December 4th

To foster international participation, this course will be held online

Bayesian methods provide a principled probabilistic framework for reasoning under uncertainty, a capability central to modern artificial intelligence (AI) and biomedical research. This course introduces the foundational frameworks of Bayesian estimation, Bayesian computation, and Bayesian inference for computational biology and biomedical data science. Over five half-days, the course combines lectures with practical exercises to give participants the skills and confidence to run common Bayesian analysis workflows independently using freely available resources. Throughout, biological datasets serve as motivating examples for hands-on exercises using publicly available tools and resources.

This course is designed for advanced MSc, PhD, and postdoctoral researchers in statistics, biostatistics, and computational biology. Working knowledge of R/Bioconductor is required. Familiarity with probability and statistics and some experience with linear and logistic regression are also expected. Prior exposure to machine learning is helpful but not required.

By the end of this course, participants will be able to:
•    Articulate the Bayesian machine learning paradigm and contrast it with non-Bayesian machine learning approaches
•    Specify appropriate prior distributions, including informative, weakly informative, and non-informative priors
•    Derive and compute posteriors for single-parameter and multi-parameter models using conjugate and non-conjugate approaches
•    Understand and implement MCMC algorithms (Gibbs sampling, Hamiltonian Monte Carlo) and variational inference, recognizing their roles as computational engines of Bayesian AI
•    Write and debug models in R/Python; diagnose convergence using trace plots, R-hat, and effective sample size
•    Apply Bayesian methods to multimodal data integration and uncertainty quantification in computational biology

Daily online sessions, 14:00–18:00 Berlin time (CET); offline communication will take place via Slack.

MONDAY, NOVEMBER 30
•    14:00–14:15: Welcome and practicalities
•    14:15–14:45: The Bayesian paradigm: priors, likelihood, and the posterior
•    14:45–15:15: Single-parameter models and conjugate priors
•    15:15–16:00: Multi-parameter models: joint and marginal posteriors
•    16:00–16:30: Break
•    16:30–17:30: Hands-on - Introduction to Bayesian paradigm
•    17:30–18:00: Q&A and discussion 

TUESDAY, DECEMBER 1
•    14:00–14:15: Recap
•    14:15–14:45: MCMC: Gibbs sampling and Hamiltonian Monte Carlo
•    14:45–15:15: Convergence diagnostics
•    15:15–16:00: Bayesian model evaluation 
•    16:00–16:30: Break
•    16:30–17:30: Hands-on - MCMC
•    17:30–18:00: Q&A and discussion 

WEDNESDAY, DECEMBER 2
•    14:00–14:15: Recap
•    14:15–14:45: Hierarchical (multilevel) models and shrinkage
•    14:45–15:15: Generalized linear models in the Bayesian framework
•    15:15–16:00: Bayesian non-linear regression
•    16:00–16:30: Break
•    16:30–17:30: Hands-on – Bayesian hierarchical GLMs and non-linear regression
•    17:30–18:00: Q&A and discussion 

THURSDAY, DECEMBER 3
•    14:00–14:15: Recap
•    14:15–14:45: Bayesian regularization
•    14:45–15:15: Variational Bayes
•    15:15–15:45: Break
•    15:45–16:45: Hands-on – Bayesian regularization 
•    16:45–17:30: Hands-on – Variational Bayes
•    17:30–18:00: Q&A and discussion 

FRIDAY, DECEMBER 4
•    14:00–14:15: Recap
•    14:15–14:45: From single-view to multiview: the Bayesian case for data integration
•    14:45–15:15: Bayesian multimodal AI paradigms: early, intermediate, and late fusion 
•    15:15–15:45: Break
•    15:45–16:15: Bayesian factor models for multimodal integration
•    16:15–16:45: Uncertainty quantification in Bayesian multimodal models
•    16:45–17:30: Hands-on - Bayesian multimodal integration 
•    17:30–18:00: Final Q&A and course wrap-up

1 - Machine Learning for Multi-Omics Integration - ONLINE, 21-23 September

2 - Machine Learning for Drug Discovery - ONLINE, 12-15 October 

3 - Causal AI Methods for Computational Biology - ONLINE, 2-6 November

4 - Deep Learning for biologists - ONLINE, 23-27 November