The course is intended primarily to Master’s degree statistics students or Ph.D. students in statistics, but it is also suitable for Master’s degree and doctoral students and exchange students from other disciplines with sufficient background in Bayesian statistical inference. The course is also suitable for the 2nd and 3rd year Bachelor degree students of statistics who have taken Bayesian inference course.
Missing data are common in many research problems, for example in health surveys. Missing data mechanisms are often selective, and simple complete-case analyses can yield biased results. This course introduces some commonly applied methods to handle missing data.
Studies in probability, statistical inference, Bayesian inference, data analysis with R.
Contents: The topics covered include
- types of missing data
- need to handle missing data
- techniques to handle different types
- multiple imputation
- data augmentation based on Bayesian inference
The course will alternate between lectures (2h/week) and exercises (2h/week).
Scope 5 cr
Tommi Härkänen, Docent