Bayesian Statistics (MAT0070)

Table of Contents

Program overview

Bayesian methods are increasingly popular in both industry and academia. The course aims at providing a modern introduction to Bayesian statistical methods, covering the fundamentals of both the parametric and the nonparametric approach. This is a 48 hours-module, and I cover the first 24 hours on parametric Bayes modelling. Here you learn about the philosophy of the Bayesian approach as well as how to implement it for common types of data. We compare the Frequentist and Bayesian approaches, and appreciate some of the benefits of the latter (e.g., uncertainty quantification, more intuitive and interpretable results). The course also comprises a 24-hours module on non-parametric Bayes taught by Prof. Ruggiero.

What you will learn (parametric Bayes)

  • Motivation and formal setting for the parametric Bayesian approach
  • Exchangeability and de Finetti’s theorem
  • One-parameter models (Binomial, Poisson, Exponential)
  • Exponential families and conjugate priors
  • Point and interval estimation
  • The Normal model and other multi-parameters models
  • Bayesian regression

Meet your instructors

Silvia Montagna
Matteo Ruggiero

FAQs

Are there prerequisites?

Stochastic Modelling for Statistical Applications (MAT0039)

How often does the course run?

Every Fall semester

Refer to the course webpage for a complete overview of the course.