Student Seminar in Statistics

Topic: Bayesian Statistics

Description:

The seminar discusses the Bayesian paradigm where also the unknown parameters are considered as random variables. Topics include prior, posterior and likelihood, differences to frequentist statistics, empirical Bayes procedures, nonparametric Bayesian methods, asymptotic properties of the posterior, model selection and computational methods.

The talks can be given in either German or English.

Literature:

Robert, Ch. (2007). The Bayesian Choice, 2nd ed., Springer NY.

For additional material please check the program.
Handouts from a course on Bayesian statistics can be obtained here.

Program:

The assignment of the talks will take place on Monday, February 18.
List with possible topics for a talk + further references.
Schedule of talks.

Synopses:

• The Bayesian Paradigm
• Inference and Decision Theory
• Prior Information
• Nonparametric Bayes: The Dirichlet Process
• Asymptotics
• Lookback: Slides Dirichlet Process, Derivation of the posterior distribution for Dirichlet Process Priors
• Model Selection
• Hierarchical and Empirical Bayes
• Computational Aspects
• Subjective Probability