Entry requirements

120 credits including 90 credits in statistics, or 120 credits including 60 credits in statistics and 30 credits in mathematics and/or computer science.

Learning outcomes

After completing the course, the student is expected to

• have knowledge of basic concepts, philosophy and views in Bayesian statistics
• be able to derive posterior distributions in simple cases
• be able to derive and use predictive distributions
• be able to identify and formulate Bayesian statistical models for analysis and prediction
• be able to formulate and estimate models with modern computer-based methods for approximation of posterior distributions
• understand and be able to use basic principles for decisions under uncertainty
• have knowledge of how to use Bayesian methods for model comparisons
• be able to independently and critically review use of Bayesian methods
• in writing and orally present a completed statistical analysis with Bayesian methods

Content

Bayes' theorem, prior distribution, posterior distribution, predictive distributions, conjugate prior distributions, basic decision theory, Bayesian model comparison, posterior approximations, MCMC, probabilistic programming

Instruction

Instruction is given in forms of lectures, exercise sessions and computer labs

Assessment

Assessment takes place through a small data analysis project at the end of the course, as well as through mandatory hand-in assignments/tests.