# Announcements:

April 21, 2009: the assignment and exam results are available.

Errors and remarks:
Slide 395 ('Making a decision'): both occurences of s_{ij} should be replace by s_{ji}, since we take the first index to indicate the row and the second to indicate the column of the matrix.
Exercise 6.3: 'status quo' does not have a special meaning in the context of decision problems, so it just means 'the current situation, nothing changes';
Exercise 6.4: the answers to part c) are incorrect. You can use one of three approaches to solving this exercise, each will result in slightly different risk premiums. 1) Draw a curve through the given points and read all necessary information from your drawing. 2) Observe that the function u(x) = sqrt(x)/1000 fits the points; you should then find risk premiums of 7996 and 113594. 3) Use linear interpolations between the points and you will find riskpremiums of 0 and 115.000.

General:
Check this page regularly, for the overview of subjects may be updated regularly.

## Contents

Goal
Organisation
Subject overview
Exam and Practical assignment
Literature and other coursematerial
Software
Course evaluation

## Goal

Acquiring knowledge and understanding of different models and decision analysis techniques for (rational) decision making with and without uncertainty. Roughly, the following subjects will be discussed:

• decision criteria
• decision trees
• utility theory
• multi attribute decision making
• analyses of uncertainties

## Organisation:

This course is given in Period 3: weeks 6 - 15 (2009); the final exam is in week 16.

Lectures are given twice a week. For further details, see the official schedule.

Lecturer:
dr. Silja Renooij (e-mail)

The following table gives an overview of the subjects treated each class. The corresponding slides can be downloaded. In addition, exercises to practice with the material are listed. Note that more in-depth understanding will be required for passing this course than necessary for the exercises.

The exercises can be found in the studymanual, which also lists the relevant chapters of the textbook (see below).

 Week Date Subjects Slides Relevant exercises 6 04-02 Introduction Slides 1 1.1,1.2 06-02 Decision criteria; Decision trees " 2.1, 2.2, 2.3, 3.1, 3.2ab, 3.3ab, 3.4, 3.5 7 11-02 Probabilities Slides 2 3.2c, 3.3c, 3.4ef, 4.1, 4.2 4.3 13-02 Lotteries " 4.4, 5.1a, 5.4 8 18-02 Utilities " 5.1b, 5.2, 5.3, 5.5, 5.6 20-02 Risk attitudes " 6.1, 6.2, 6.3, 6.4, 6.6 9 25-02 Multi-attribute utility theory Slides 3 6.5, 7.2, 7.16 27-02 MAUT cont'd " 7.1, 7.3, 7.7, 7.8 10DEADLINE! 04-03 MAUT cont'd " 7.4, 7.5, 7.6, 06-03 MAUT cont'd Slides 4 7.9, 7.17 11 11-03 MAUT in practice " 7.10 -- 7.15 13-03 MAUT in practice cont'd " 12 No classes (re-examinations) 13 25-03 Sensitivity analysis Slides 5 8.1, 8.2 27-03 Sensitivity analysis cont'd " 8.3, 8.7a-c 14 01-04 Value of information analysis " 8.4, 8.5, 8.6, 8.7de 03-04 Analytic Hierarchy process " 9.1 -- 9.7 15DEADLINE! No lectures: Study week 16 Exam

The course will be graded based on a practical assignment and a written exam.

• Both these tests will be marked on a scale from 1.0 to 10.0 (for interpretation, see the Dutch grading system);
• to receive a final grade (FG), both the assignment grade (AG) and the grade for the written exam (EG) should be at least a 5.0 (see re-examination conditions below);
• the final grade is then determined as follows:

FG = Mark(1/3 AG + 2/3 EG),

where Mark(x) =

= Floor(x), if x < 5.8
= Round-to-nearest-half(x), otherwise

(Also see Art. 5.3 – marks of the Education and Examination Regulations Graduate School of Natural Sciences:
"The results of final tests are expressed in numbers from 1 to 10; in half numbers from 6 to 10 and in whole numbers below 6.
The final assessment of a course is either pass or fail, expressed in numbers: 6 or higher and 5 or lower, respectively.")

To qualify for the 2nd chance exam in week 23, you should have received at least a 5.0 for the practical assignment in this year. Under certain conditions, you may be granted a second chance for the practical assignment. It is not possible to redo both the written exam and the practical assignment! See the following table for details:

 Re-examination conditions for the  written exam  : Re-examination conditions for the practical assignment  : the exam is taken in week 23 (2009). a report on the practical assignment was submitted in period 3 of 2008/2009, and graded with at least a 5.0 a report on the practical assignment was submitted in period 3 of 2008/2009, and the assignment grade was less than 5.0, and the exam grade was at least 5.0, and a different subject is chosen, and an appointment is made with the lecturer concerning subject and timeschedule.

Partial results from previous years are no longer valid. Reports on practical assignments done in previous years may be re-submitted, but are not guaranteed to receive the same grade due to changes in criteria.

## Exam and practical assignment

The final written exam is a closed book exam. You are strongly advised to bring a calculator (mobile phones are not allowed!)

Here's a list of what I expect you to know and understand for the exam. Here's an example of a written exam. Note that this exam stems from the time when participants could consult the book, notes, and transparencies during the exam; this is no longer allowed!

The practical assignment consists of performing a decision analysis and writing a report about it. The assignment may be done either alone or, preferably, in pairs. View the grading criteria based on which your report will be marked.

• Monday 2 March 2009, 17:00 : e-mail to Silja Renooij describing in plain text (ASCII) your choice of decision problem;
• Monday 6 April 2009, 17:00 : 10 page written report, on paper, to be submitted to Silja Renooij (pigeon hole, CGN, B205).
Important notes:
1. Deadlines will be strictly enforced.
2. Submissions received after a deadline will be disregarded; you can then no longer pass the course and do not qualify for re-examination.

## Literature and other course material

Book:
Robert T. Clemen, Terence Reilly (2004)
Making Hard Decisions -- With DecisionTools
Duxbury Press. ISBN 0495015083

The 2001 edition of this book can also be used (ISBN 0534365973);
the only difference with the above book is an update of the accompanying software. (see also the book's website)

The above book has a blue cover. An even older version of the book, with a green cover, does not include software,
but does contain all relevant text. It can therefore be used for the course as well:

Robert T. Clemen (1996)
Making Hard Decisions -- An Introduction to Decision Analysis
Duxbury Press, ISBN 0534260349 (also see its website)

It is entirely your own responsibility to see to it that you have one of the books in time for the course.

Slides:  (see overview) the slides are considered essential supplementary material. A few topics discussed in the course are not covered by the book; in addition, a number of topics are covered in more detail than provided by the book. All relevant information can be found on the slides.

R.L. Keeney & H. Raiffa, Decisions with Multiple Objectives, Cambridge University Press, 1993
chapters of interest:  Ch 4 (utility theory), 5,  6 (Multi Attribute Utility Theory)

W.L. Winston, Operations Research: Applications an Algorithms, Duxbury Press, 1994
chapters of interest: Ch 13 (decision criteria),  14 (MAUT,  Analytic Hierarchy Process)

J.A. Rice,  Mathematical Statistics and Data Analysis, 2nd edition, Duxbury Press, 1995
chapters of interest: Ch 15

H.C. Sox, M.A. Blatt, M.C. Higgins, K.I. Marton (1988), Medical Decision Making, Butterworth-Heinemann, 1988.

## Software

To enhance understanding of different subjects, we recommend trying out the demo version of, for example, DPL. More information about decision analysis in general and software more in particular can be found on the website of the Decision Analysis Society (DAS).

## Course evaluation

To guard the quality of our educational programme, each year every course is evaluated by the students that registered for the course. The student-evaluation helps instructors to evaluate and if necessary adapt their courses and allows the Institute to evaluate its teachers.  For the evaluations to be useful it is necessary to get respons from a large number of participants.

Our intention is that you find the decision analysis course interesting, motivating, perhaps a bit difficult now and again but doable, challenging and enjoyable, and last but not least, we hope that you feel that you have learnt many useful things. If for some reason the course wasn't what you'd expected and/or if you have any (reasonable) suggestions to improve the course and everything that comes with it, please fill out the evaluation form. If you're completely satisfied then we'd also like to know that and filling out the  evaluation form would be a way of telling me this.

For those who are interested, links are provided to the evaluation of previous years. Note that this can be viewed internally only. For each evaluation, you'll find listed the number of registrants (union of OSIRIS subscribers and participants), the number of participants (distinct students that participated in an examination in that year), and the number of respondents (students that filled out the evaluation form).

Please note that the evaluation site contains links to evaluations of a 'Besliskunde' course in years previous to 2002-2003. This is a different course from the current one as it was a third year course, entirely focussed on medical decision making; evaluations are therefore incomparable.

• 2002/2003  (36 registrants, 32 participants, 5 respondents)
• 2003/2004; period 3 "Besliskunde"  (53 registrants, 49 participants, 23 respondents)
• 2004/2005: selfstudy -- not evaluated
• 2005/2006; period 3 "Besliskunde"  (11 registrants, 8 participants, 5 respondents)
• 2006/2007; period 3 "Besliskunde"  (13 registrants, 13 participants, 3 respondents)
• 2007/2008; period 3 "Besliskunde"  (16 registrants, 14 participants, 5 respondents)
• 2008/2009; period 3 "Besliskunde"  (9 registrants, ?? participants, ?? respondents)

22 January 2009.