S-114.600 Introduction to Bayesian Modeling
Fall 2003 (2 cr)
27+27 (2+2) f
Teachers: Dr.Tech. Aki Vehtari, M.Sc. Toni Tamminen
Bayesian probability theory and Bayesian inference.
Bayesian models and their analysis. Computational methods, Markov chain
Requirements:Exam and exercises.
Literature: Gelman, Carlin, Stern & Rubin: Bayesian Data
Analysis, Second edition, and other material announced on course web
Prerequisites: First year mathematics are recommended.
Lectures: Mondays 12-14, room E111, in Finnish.
Computer exercises: Tuesdays 14-16, computer lab F402
E111 and F402 are in the Electrical and Communications
Engineering building, Otakaari 5.
Exam: 9.12.2003 9-12 hall S4
Lectures are in Finnish, but it is possible to take this course in
English by self studying the relevant parts of the book. Instruction
in computer exercises is given mainly in Finnish, but it is possible
to get some personal instruction in English. The assistant in the computer
exercises is MSc Toni
Here is the list of chapters and pages studied in this course. For
each chapter there is also a list of exercises. The assistant will help
with these exercises in the computer lab. Some of the exercises are
marked with *. The assistant will also help with these exercises in
the computer lab, but you need to write a report with your solution,
results and discussion. More details below.
There is a hint page for some of the exercises.
- Background, Ch 1: 3-27, Ex: 1, 2, 4, 6, 7.
- Single-parameter models 1, Ch 2: 33-45, Ex: 1, 2, 3, 4, 9.
- Single-parameter models 2, Ch 2: 46-55, 61-65, Ex: 8, 11*, 18.
- Introduction to multiparameter models, Ch 3: 73-94, Ex: 2, 3*, 4*, 6.
- Large-sample inference and frequency properties of Bayesian inference
, Ch 4: 101-112, Ex: 1*
- Hierarchical models, Ch 5: 117-127, 131-150, Ex: 1*, 2, 6
- Introduction to computation, Ch 10, 11, 12: 275-282, 283-308, 311-312, Ex:
10.1, 11.3*, Metropolis and
Gibbs sampling**, BUGS*
- Decision analysis, Ch 22: 541-544, 552-555, 567-568, Ex 1*
- Model checking and improvement, 6: 157-189, Ex: 1*, 7, speed of light
- Modeling accounting for data collection, Ch 7: 197-237, Ex: 1
Final grade of the course is the average of the exam and the exercise report
provided that you pass (get a '1') both of them.
You are allowed to take one A4 page of notes with you to the exam. You
may write anything you think might be helpful as long as you have
only one A4 paper. This way you don't need to try to remember
everything but instead can spend more time learning and understanding things.
Because there is quite lot of text in course book, here is a
more detailed list of things which are required in exam. Exam hints.
You need to write an exercise report for the exercises marked with
*. Describe your solution briefly, show the code you have used,
and describe and discuss your results. You will get 1
point for the results and additional 1 point for
meaningful discussion. In discussion you may write about which
parts were unclear or what questions or ideas did this exercise
raise. You may also discuss the relevance of the exercise for
The maximum number of points is 11x2=22, and the grade of the
report is decided based on following: 12-13=1, 14-15=2, 16-17=3,
18-19=4, 20-22=5. This means that you don't need to solve all the
*-exercises to pass the course.
Deadline of this report is 25.11.
You are allowed to write the report in groups of two or three.
Return just one report with the names of all students in the group.
S-114.600 Introduction to Bayesian Modeling is not considered a post-graduate course.
In Fall 2003 post-graduate students may take the course with the code
Post-graduate students will get 2 credits.
- S-114.230 Yksilöllisiä laskennallisen tekniikan opintoja (1-6 ov) L
- Bayes, 1763: An Essay Towards
Solving a Problem in the Doctrine of Chances
(Reprint available at JSTOR)
- Stigler, 1986: Thomas Bayes's Bayesian Inference
(Available at JSTOR)
You may send your comments to the lecturer
Helsinki University of Technology
Lab. of Computational Engineering
P.O.Box 9203, 02015 HUT
tel: 451 4849
This page is maintained by
This page has been updated 5.12.2003