Staff: Professor Trevor Bailey, College of Engineering, Mathematics and Physical Sciences, University of Exeter, UK.
This course is designed to provide an introduction to Bayesian approaches to statistical modelling and associated methods (such as Markov Chain Monte Carlo (MCMC) techniques and variants of that. The focus will be very much on practical implementation of the methods using the statistical programming language R and related packages (including links between R and OpenBUGS).
The course is arranged into sessions:
- Introduction to Bayesian Modelling and MCMC
Review of the Bayesian Approach, MCMC and Hastings-Metropolis algorithm, model checking and inferential ideas.
- Common Applications of Bayesian Modelling
Examples of implementing a range of standard models using R and links between R and OpenBUGS, including Linear and Generalised Linear Mixed models, survival models, temporal and spatio-temporal models, Generalised Additive models.
- Practical Workshop
An opportunity for groups of participants to work through modelling examples and associated computer implementation for themselves.
- Extensions and more Advanced Ideas
A brief discussion of more advanced techniques, including Reversible Jump (RJMCMC), and Approximate Bayesian Computation (ABC).
The course is appropriate for those involved in statistical analysis in application areas such as epidemiology, public health, medical statistics, environmental science and bioscience. It will assume background knowledge of routine statistical ideas such as regression and generalised linear models and related concepts and some familiarity with R. It will not assume a high level of theoretical mathematical statistical knowledge.
Professor Trevor Bailey, College of Engineering, Mathematics and Physical Sciences, University of Exeter, UK.
Registration deadline - April 6th
Course held in english.