Bayesian Modelling

Bayesian modelling is a free course for University of St Andrews staff and research degree students. The course covers applied Bayesian inference to some common statistical inference problems such as linear and non-linear regressions and classifications. Some Bayesian machine learning concepts will also be covered.

The methods covered will be implemented using a new programming language, Julia, and the course includes a substantial practical component.

The course is facilitated by St Leonard's Postgraduate College and CREEM and delivered by the School of Computer Science.

The Course Material

https://lf28.github.io/BayesianModelling/

Content and Structure

The course covers three main parts:

1. The modelling principles that underlie the Bayesian statistical paradigm
   • The modelling principles of Bayesian inference
   • Directed graphical models as a modelling tool
2. Approximate Bayesian inference algorithms
   • MCMC algorithms
3. Applied Bayesian modelling with probabilistic programming languages
   • Regression problem
   • Classification problem
   • Generalised regression problem

Learning Outcomes

After taking the course, participants should be able to:

• Understand the principles that underline the Bayesian statistical paradigm
• Understand the needs of the Bayesian statistical paradigm in machine learning and statistical learning
• Understand the main computational algorithms for implementing Bayesian statistical inference
• Use probabilistic programming languages to do applied Bayesian modelling and computation

Book Your Place

Enrol on the course through the Personal Development Management System (PDMS).