Dr Sara Geneletti Inchauste Col 5.07

This course is available on the MPA in Data Science for Public Policy, MSc in Data Science, MSc in Health Data Science, MSc in Statistics, MSc in Statistics (Financial Statistics), MSc in Statistics (Financial Statistics) (Research), MSc in Statistics (Research), MSc in Statistics (Social Statistics) and MSc in Statistics (Social Statistics) (Research). This course is available with permission as an outside option to students on other programmes where regulations permit.

The course will require the use of computers so students must have a laptop.

Students must have completed Elementary Statistical Theory (ST102).

Basic knowledge in probability and a first course in statistics such as ST102 or equivalent probability distribution theory and inference. Basic knowledge R or an equivalent programming language required. Students who do not have prior knowledge of R will be required to take an R module with the Digital Skills Lab.

The course is a hands-on introduction and development of the analysis of Bayesian models with using  Nimble (similar to BUGS), Stan and R-INLA with focus on data sets and application. The main topics will be Bayesian regression, Bayesian Hierarchial models and Spatial models. Throughout the course there will be practical examples from epidemiology, public health and social science which will involve data analysis.

This course will be delivered using a combination of lectures, seminars and Q&A sessions totalling a minimum of 33 hours in the Winter Term. Week 6 will be used as a reading week.

The course will cover the following

Spatio-temporal data: what is it and why is it useful?

Basics of the R programming language

Bayesian methods, with focus on an intuitive understanding and practical data analysis.

• Bayesian foundations
• Markov Chain Monte Carlo 
• Regression (including GLMs)
• Hierarchical Models
• Hamiltonian Monte Carlo
• Integrated Nested Laplace Models
• Spatial models including areal data, ecological regression and spatial prediction

Bayesian software 

• Nimble (similar to BUGS)
• Stan
• R-INLA
• Discussion of advantages and differences between software

There will regular workshops where students have to replicate the analyses covered in lectures in different data sets

Data analysis and regression using multilevel/hierarchical models: Andrew Gelman and Jennifer Hill

Statistical Rethinking: A Bayesian Course with Examples in R and STAN

Spatial and Spatio-temporal Bayesian models with R-INLA: Marta Blangiardo and Michela Cameletti

Project (20%, 2500 words) in the WT Week 6.
Presentation (30%, 1500 words) in the WT Week 11.
Project (50%, 6000 words) in the ST Week 2.