Transcript Decision analysis and risk management: Introduction to course

Decision analysis and risk management:
Introduction to course
Jouni Tuomisto, THL
Outline for session 1 (28.2.)
• Some basic concepts of decision analysis
• Introduction and objectives of the Darm course
• Practicalities: schedule, website
– http://en.opasnet.org/w/Darm
• Group work: decision analysis on swine flu
• Individual work: risk management analysis on swine
flu and narcolepsy
• Evaluation and credits
• Intro to the swine flu case in Finland
Outline for session 1 (28.2.)
Introduction and objectives of the DARM course
Practicalities: schedule, website
http://en.opasnet.org/w/Darm
Lectures and classroom exercises
Case study exercise
Group work: decision analysis on swine flu case
Individual work: risk management analysis on swine flu
Evaluation and credits
Guidance, help, and communication
What are DA and RM?
Introduction to the swine flu case in Finland
Probability
• In this course, we take the subjective interpretation
of probability:
• Probability is an individual's degree of belief in a
statement, given the evidence.
• → Everyone has his/own probability.
• → A person's probability about something may
change in time and when the evidence changes.
Probability: a standard
• There are N balls in an urn. They are otherwise
similar except that R are red and the rest are white.
• One ball is picked at random (random = in a way
you believe that each ball is equally likely to be
picked).
• What is the probability that a red ball is picked?
Probability of a red ball
• p(x|K) = R/N,
– x=event that a red ball is picked
– K=your knowledge about the situation
Probability of an event
• What is your probability that a bus arriving at Kuopio
marketplace will be at time (=late less than 5
minutes)?
• We could try to get bus statistics, but there is no
time for that. We need the probability now.
• How to proceed?
Probability of an event x
p
Decision 1
1-p
Red ball
White ball
Prize
100 €
0€
x happens 100 €
Decision 2
x does not
happen
0€
• If you are indifferent between decisions 1 and 2,
then your probability of x is p=R/N.
Decisions under uncertainty: Trip
mode
• You are going from point A to a meeting near the
Kuopio market place. Should you take your car or
the bus?
• You have a monthly ticket for the bus, so there is no
extra cost there.
• However, the bus may be late and the meeting may
start without you.
• What are the possible outcomes?
What are the possible outcomes?
• Take the bus, be in time.
• Take your car, be in time.
• Take the bus, be late.
• This is the order of preference, but how much better
or worse are they in comparison?
• How to quantify preferences?
Utility as a measure of preference
• Take the bus, be in time. u=1
• Take your car, be in time. u=?
• Take the bus, be late. u=0
• The u(best outcome)=1, u(worst outcome)= 0.
• How to assess those in between?
Utility of an event x (Car, in time) Utility
Car, in time
?
Bus, in time
1
Bus, late
0
Decision 1
pt
Decision 2
1-pt
• Adjust p in such a way that you are
indifferent between decisions 1 and 2.
• Then, your utility u(x)=pt.
Decision analysis of trip mode choice
Utility E(u)
Car, in time
u
u
Decision 1
pt
Decision 2
1-pt
Bus, in time
Bus, late
1
0
• Calculate your expected utility for each decision and
choose the highest.
• E(u(D1,x))=u*1; E(u(D2,x))=1*pt+0*(1-pt)=pt
pt
Decisions by an individual vs. in a
society
• In theory, decision analysis is straightforward with a single
decision-maker: she just has to assess her subjective
probabilities and utilities and maximize expected utility.
• In practice, there are severe problems: assessing probabilities
and utilities is difficult.
• However, in a society things become even more complicated:
–
–
–
–
There are several participants in decision-making.
There is disagreement about probabilities and utilities.
The decision models used are different.
The knowledge bases are different. NOTE! In this course,
"knowledge" means both scientific (what is?) and ethical (what
should be?) knowledge.
Societal decision example:
bus transport in Kuopio
• Buses are often late, thus many inhabitants in
Kuopio prefer cars.
• Should the city subsidise to improve bus service
and thus increase bus trips?
– Increase in living conditions.
– Less pollutants.
– More attractive city.
– BUT: It costs money.
– Actions may be ineffective etc.
Bus transport subsidies in Kuopio?
• Who defines the problem? Whose utilities?
p(in time), cost #trips E(u)
Decision:
BAU
pt
p(# trips increase)
?
?
p(improve)
?
Decision: subsidise
with 100 k€
?
0€
?
100 k€ ?
100 k€ ?
100 k€ ?
100 k€?
?
?
Theoretical solution:
everyone can participate
• Risk management and decision analysis methods
should allow for this.
Wiki pages with questions and answers are
used to organise information needed
Statements and argumentation are
used to organise different opinions.
Objectives of the course.
The student should:
•
•
•
•
Get a good overview of modern assessment methods.
Learn basic concepts of decision analysis.
Learn to use decision analysis in practice.
Understand the connections between societal decisionmaking and decision analysis.
• Be able to apply the scientific method and falsification in the
context of decision analysis.
• Know how to build a decision analysis based on the
requirements of risk management.
• Be able to utilize modern web workspaces for decision
analysis.
Schedule of the course
– http://en.opasnet.org/w/Darm
Lectures and classroom exercises
Theory of DA & RM
Swine flu story discussions
Considering the theory in a practical context
Exercises
Calculation exercises
Using Opasnet
Case study exercise
Analyze the swine flu case and consider yourself as
giving advice to the Ministry of Social and Health
affairs of Finland
What could be learned from the swine flu case for
improving public health risk management in Finland?
Exercise consists of two parts:
Decision analysis study plan (group work)
Risk management analysis (individual work)
Case study exercise description and instructions:
http://en.opasnet.org/w/Category:DARM_exercise
Decision analysis study plan
Plan a decision analysis study on the swine flu case
Background description
Purpose and scope of the study
Analysis plan
(Expected) results
If possible, the plans can try to be realized, at least partially
Write the plan in Opasnet
Group work
3-4 people/group
Recommended: at least one fluent in Finnish in each group
Some swine flu case material only available in Finnish
Risk management analysis
Analyze risk management in the swine flu case and
consider alternative ways to making it
Background (cf. DA study plan background)
RM in the swine flu case
Alternative approach to (a part of) RM in the swine flu
Alternative vs. actual
Conclusions and recommendation (for the Ministry)
Write your report in Opasnet
Individual work
Not limited by the scope of the DA study plan of the
group one attended
Hints to making case study exercise
Build on:
theory lectures and classroom exercises on this course
classroom discussions on the swine flu case as a DA and RM
problem
materials listed and linked to on the course web-page
the demonstrator DA model
assessments in Opasnet
descriptions of assessment and variable objects in Opasnet
other related information e.g. on the web and libraries
your own expertise and opinions
other groups'/individuals' exercise works
Clear and focused scoping is important!
Using Opasnet
Basics: 4.3. 9-12, MC9
Possible problems can be discussed in classroom as
they occur
Discussion and argumentation exercise 1.4. 9-12,
MC9
Evaluation and credits
DARM: 6 ECTS
Whole course obligatory for ToxEn students
For others partial accomplishments are also possible
Lectures and classroom exercises: 3 ECTS
Case study exercises: 3 ECTS
Decision analysis study plan 2 ECTS
Risk management analysis 1 ECTS
Lectures are voluntary, but highly recommended
Active participation practically necessary for successful
accomplishment of exercises
Evaluation and credits
2/3 * exercise grade + 1/3 * classroom exercise grade
= overall course grade
Evaluation and credits
Classroom exercise evaluation
Scoring of some calculation exercises
Marko will explain details at the exercises
Bonus for active participation in classroom
Evaluation and credits
Exercise evaluation
DA study plan = 2/3 of the exercise grade
RM analysis = 1/3 of the exercise grade
Bonus for active commenting and discussion in
Opasnet
Clarity and reasoning (not length)
Basis for evaluation (to be) explained in more detail on
the case study exercise page
Guidance, help and communication
during the course
General arrangements
Case study exercise
Technical help
Commenting and discussion in Opasnet
End
What are DA and RM?
Swine flu case in Finland
What is probability?
–
1. Frequentists talk about probabilities only
when dealing with experiments that are random
and well-defined. The probability of a random
event denotes the relative frequency of
occurrence of an experiment's outcome, when
repeating the experiment. Frequentists consider
probability to be the relative frequency "in the
long run" of outcomes.[1]
– 2. Bayesians, however, assign probabilities to
any statement whatsoever, even when no
random process is involved. Probability, for a
Bayesian, is a way to represent an individual's
degree of belief in a statement, or an objective
degree of rational belief, given the evidence.
– Source: Wikipedia