Advanced Bayesian Modelling with BUGS
This course is designed for statisticians who want to improve their command of the BUGS modelling software. We will assume a previous knowledge of BUGS to the level of our companion course Introduction to Bayesian statistics using BUGS, or equivalently the first few chapters of The BUGS Book. Through a mixture of lectures and practical exercises, participants will learn how the BUGS language can be used to build models that represent the complexities of real data.
The OpenBUGS, WinBUGS and JAGS software and their R interfaces will all be covered.
- Recap of Bayesian inference and principles of BUGS: graphical models.
- Hierarchical models, meta-analysis and prior distributions.
- Missing data.
- Censoring, truncation and defining new distributions.
- Complex hierarchical models, model checking and robustness.
- Evidence synthesis: network meta-analysis and general synthesis.
- Measurement error.
- Model checking in evidence synthesis: conflict and bias.
Throughout we emphasise thoughtful choices of prior distribution, assessing the influence of model assumptions, and checking and comparing fitted models.
Much of the course material is based on The BUGS Book (Lunn et al., 2013), which is provided with the course fee.
Statisticians working in any application area who are familiar with Bayesian modelling and the BUGS or JAGS software.
Participants should bring a laptop with OpenBUGS or JAGS installed, and (optionally) the R2OpenBUGS or rjags R packages.