Skip to content

Applied Bayesian Data Analysis

Published: 4 May 2017

Course syllabus

Course name: Applied Bayesian data analysis

Credits: 7,5 Credits

Education Level: Graduate Level

Syllabus approved by:

Examiner: Professor Inge Söderkvist

Prerequisites: Knowledge in probability and statistics equivalent undergraduate introductory course S0001M or S0008M on technical faculty. Programming experience in e.g. Julia, python, MATLAB, R or similar languages.

Objectives/Expected outcomes: After the course the student will be able to know and apply standard Bayesian models that is covered in the course on the various types of data. (S)he will be able to interpret the results from the analysis and validate the model’s ability to describes the observed data.

  1. Knowledge and understanding

    1. Know the basic Bayesian statistical concepts.

    2. Know and identify the most common pitfalls associated with the models, the calculations and the methods given in the course.

  2. Competence and skills

    1. Design Bayesian statistical models for different types of experiments and data types.

    2. Be able to apply sampling methods and check the convergence.

    3. Interpret the posterior.

  3. Judgement and approach

    1. Be able to assess when a particular type of Bayesian model and method is applicable.

    2. Describe and justify the choice of the prior distribution.

    3. Validate how well the model describes data with known model validation techniques.

Course content: The course covers various methods in Bayesian data analysis

which can be used to make conclusions from empirical data in combination with prior experience and knowledge. The main reasons for applying Bayesian analysis are: (i) to avoid nonidentiability problems (e.g. due to overparameterization, separation and colinearity); (ii) to provide richer inference and avoid the common problems associated with p-values (e.g. dependencies on the sampling and testing intentions); (iii) to make direct probability statements about the parameter you are interested in given the observed evidence (e.g. the parameter value is, given the observed data, inside the credible interval with a measurable probability).

The course covers and applies the common Bayesian models on various types of data, experiments and examples. We look at cases with no prior information, with much prior information and the hierarchical case where we let the data inform us about our prior distributions.

Realization: Instruction consists of lectures and seminars. Compulsory assignments are done continuously during the course. The course requires active participation in the seminars where individual solutions are presented and defended.

Examination: The course will be examined based on individual assignments and your presentation at the seminars.

  • Written course assignments 5.5 credits

  • Oral presentations/course assignments 2 credits

Grading: G or U

Course: LP1-LP2, 2019.

Application to: Jesper Martinsson (

Apply before: 2019-06-15.

Literature: John K. Kruschke (2014). Doing Bayesian Data Analysis: A tutorial with R, JAGS, and Stan. Elsevier Science, 2nd edition.

Contact: Jesper Martinsson (