Bayesian Data Analysis

Classical statistics offers a powerful toolbox for data analysis. This toolbox, however, may not always be sufficiently flexible for modern data situations. For example, some situations benefit from data integration or the inclusion of information from other sources that your data. The Bayesian framework allows for the integration and inclusion of information from many sources as well as a natural quantification of uncertainty in subsequent analysis. It offers these benefits for standard statistical models as well as highly customized models. This flexibility is the reason why the machinery of Bayesian inference has been successfully used in, for example, code-cracking, self-driving cars, genomic prediction, and climate-change prediction. Bayesian inference now underlies many advances in artificial intelligence, machine learning, and data science. This course offers a hands-on, example-driven approach to teaching the core concepts and tools of Bayesian data analysis.

• Explain the differences between classical and Bayesian analysis of data

• Recognize questions and situations that ask for a Bayesian approach to data analysis

• Apply modern computational approaches to Bayesian data analysis

• Effectively set up a Bayesian data analysis lifecycle

• Effectively communicate a Bayesian data analysis lifecycle

• Assignment other (70%) There are 3 assignments weighing 10, 20 and 40 % respectively. Weighted average should be 5.5
• Oral test (30%) Individual oral exam (fully protocolized)

• MAT20306 Advanced Statistics, MAT24306 Advanced Statistics for Nutritionists, MAT22306 Quantitative Research Methodology and Statistics or equivalent;
• working knowledge of R and RStudio.

• Lambert, B. (2018). A Student's Guide to Bayesian Statistics. London: SAGE. 498p.