**May 04, 2020 at 09:30 - May 08, 2020**
- BCAM

**Marco Grzegorczyk (University of Groningen, Netherlands)**

**DATES:** 4-8 May 2020 (2 sessions)

**TIME:** 9:30 - 11:30 (a total of 10 hours)

**LOCATION:** BCAM Seminar room

**ABSTRACT:**
Bayesian networks have become a popular modelling tool for learning the dependencies between variables and for representing the inferred relationships graphically in form of a network. Bayesian networks are, for example, widely applied in systems biology, where one of the most important objectives is to learn the structures of gene regulatory networks and protein activation pathways from cellular data.

The short course will focus on learning network structures from static data, i.e. independent (non-temporal) observations.

As Bayesian networks can be seen as a marriage between graph theory and statistical modelling, the course will provide an introduction to the required graph theoretic concepts [1], before the focus will turn to probabilistic modelling of graphs [2-4]. In the course, a Bayesian modelling average approach for Gaussian Bayesian networks [4] will be followed [5], so that the most important fundamentals of Bayesian Statistics, such as Markov Chain Monte Carlo simulations, will also be repeated. Finally, in the last session there will be a short R tutorial on how to apply Gaussian Bayesian networks to data.

**PROGRAMME:**
1. Graph Theory 1: Introduction (DAGs, Markov property, and the concept of d-separation)

2. Graph Theory 2: Equivalence classes of graphs (CPDAGs) and single-edge operations

3. Gaussian Bayesian networks (Bayesian modelling and the BGe score)

4. Graph inference based on Markov Chain Monte Carlo (MCMC) simulations

5. MCMC continued and short tutorial in R

**REFERENCES:**
• Chickering, M. (2002): Learning Equivalence Classes of Bayesian-Network Structures. Journal of Machine Learning Research 2:445-498

• Madigan, D. and York, J. (1995): Bayesian graphical models for discrete data. International Statistical Review, 63:215–232.

• Giudici, P. and Castelo, R. (2003). Improving Markov chain Monte Carlo model search for data mining. Machine Learning, 50:127–158.

• Geiger, D. and Heckerman, D. (2002). Parameter priors for directed acyclic graphical models and the characterization of several probability distributions. Annals of Statistics, 30(5):1412-1440.

• Grzegorczyk, M. (2010): An Introduction to Gaussian Bayesian Networks. In: Yan Qing (Ed.): Systems Biology in Drug Discovery and Development (Springer Series: Methods and Protocols: Methods in Molecular Biology, Vol. 662). Humana Press. (ISBN 978-1-60761-799-0)

** *Registration is free, but mandatory before May 1st.** To sign-up go to

https://forms.gle/snuMMuju9ywCBJUi6 and fill the registration form.

**Student grants are available. **Please, let us know if you need support for travel and accommodation expenses in the previous form before

**April 6th**.