July 3rd, (prior to the 31th International Workshop on Statistical Modelling, Rennes, France
Presenter:
Søren Højsgaard, Department of Mathematical Sciences, Aalborg University, Denmark
Goals:
Introduce participants to working with Graphical Models (GMs) and Bayesian Networks (BNs) in R. This includes probability propagation in BNs and aspects of learning BNs from data using GMs.
Topics will include:
Probability propagation with Bayesian networks (BNs) and their implementation in the gRain package.
A look under the hood of BNs to understand mechanisms of probability propagation.
Dependency graphs and conditional independence restrictions.
Learning BNs from data using graphical log-linear models in the gRim package.
Examples from genetics will be used throughout for illustrative purposes. Moreover, there will be a running example about building a BN for a medical diagnosis from real-world data.
Prerequisites:
Attendees are assumed to have a working understanding of log-linear models for contingency tables.
Preparing for the course:
Bring a laptop with the most recent version of R installed. Moreover you should install the necessary packages as:
Go to http://people.math.aau.dk/~sorenh/software/gR/ for installation instructions.
Course material
Slides/notes are available here:
Literature:
Højsgaard, S.; Edwards, D.; Lauritzen, S. (2012): Graphical models with R, Springer.