Bayesian Inference, 5 credits

Course coordinator

Silvelyn Zwanzig

Time

2024, week 7-15. Find schedule below.

Goals

The goal of this lecture course is to teach the fundamentals of Bayesian statis- tical inference. It covers basic theory, supplemented with methodological and computational aspects.

Litterature

The course is based on the textbook: S. Zwanzig and R. Ahmad: Bayesian Inference , CRC Press 2024 (in production). The registered students can download lecture notes related to the book.

Lectures

The course consists of 14 lectures + project presentation. Note that, the following list of content is preliminary. It is possible to adapt the course to research projects of the PhD students.

  1. Bayesian modelling (1 lecture, week 7), key words: Bayesian inference principle, advantages
  2. Choice of prior (3 lectures, week 7), key words: subjective and objective priors, conjugate, non{informative, reference priors
  3. Decision theory (3 lectures, week 8), key words: Bayes decision rules, minimax criterion, worst case prior
  4. Asymptotic theory (2 lectures, week 10), key words: Consistency, Schwarz Theorem
  5. Linear model (1 lecture, week 10) key words: explicite formularies for posteriors
  6. Bayes methods (2 lectures, week 11), key words: regularized estimators, HPD regions, Bayes Factors
  7. Computer intensive methods (2 lectures, week 11), key words: Gibbs sampling, MCMC, ABC
  8. Project presentations (week 15)

Examination

2 Obligatory home works

Team work is recommended

  • Assignment 1 (Bayes Modelling and Decision Theory) published published 21/2, deadline week 1/3
  • Assignment 2 (Bayes Methods) published 1/3, deadline week 12/3

Mini-Project

Application of a Bayes model to an own chosen data set. Proposal (max one page) deadline week 11, 12/3, with feedback one day later, presentation (20 min) in week 12, 20/3 10.15-12.00, written summary (max 3 pages) deadline week 15. Team work is recommended.

Oral examination or alternatively Personal home exam

Week 15. The students have the choice.


Organization

We will use the webside "studium". PH-D students has to be registered man-
ually in studium. Please send a email to me: zwanzig@math.uu.se The exact
schedule will be published next week, examination times by personal agreement
but no later than two weeks after presentation.

Schedule

Week 7

Day

Date

Time

Place

Mon

12/2

13.15-15.00

˚Ang 1167

Tue

13/2

10.15-12.00

˚Ang 4005

Wed

14/2

13.15-15.00

˚Ang 1167

Week 8

Day

Date

Time

Place

Tue

20/2

13.15-15.00

˚Ang 2004

Wed

21/2

13.15-15.00

˚Ang 1167

Thu

22/2

13.15-15.00

˚Ang 2004

Fri

23/2

13.15-15.00

˚Ang 2004

Week 9

Day

Date

Time

Place

Wed

28/2

13.15-15.00

˚Ang 2004

Thu

29/2

13.15-15.00

˚Ang 1167

Fri

1/3

13.15-15.00

˚Ang 1167

Week 11

Day

Date

Time

Place

Tue

12/3

10.15-12.00

˚Ang 1167

Wed

13/3

13.15-15.00

˚Ang 1167

Thu

14/3

13.15-15.00

˚Ang 4003

Fri

15/3

10.15-12.00

˚Ang 4003

Week 12

Wed

12/3

10.15-12.00

˚Ang 1167

Welcome!

 

FOLLOW UPPSALA UNIVERSITY ON

Uppsala University on Facebook
Uppsala University on Instagram
Uppsala University on Twitter
Uppsala University on Youtube
Uppsala University on Linkedin