This course provides a short introduction to Bayesian inference. By the end of the course, the participant should:
- Know what probability distributions are and why they are used in modelling.
- Understand the goal of statistical inference.
- Appreciate how Bayesian and frequentist approaches to inference achieve this goal.
- Know the elements required to do Bayesian inference and appreciate how they affect inferences.
- Know why exact Bayesian inference is hard.
- See how conjugate priors provide a slight remedy.
The course consists of a lecture and problem sets:
- Disease prevalence exercise. This example mirrors the material in the lectures and invites participants to estimate the prevalence of a disease. It goes through maximum likelihood estimation and Bayesian inference. The answers (written in R) to this problem set are here.
- Epileptic seizure exercise. This example uses real data from a study of epilepsy (see this paper for more information). The answers to the problem set are here.
- Need an up and running set of Python answers as well as R (I've started this but need to finish)
- Students don't understand probability distributions; probably worth including something about these in here
- May want to include the breast cancer example as an intro to Bayes' rule