Professional Documents
Culture Documents
UNDER UNCERTAINTY
• This unit:
– Discusses the
techniques to deal
with uncertainties
1
School of the Built Environment
INTRODUCTION
• Few decisions in construction industry are made with
certainty.
2
School of the Built Environment
TRADITIONAL STRATEGY
• A decision problem is characterized by decision
alternatives, states of nature, and resulting payoffs.
3
School of the Built Environment
PAYOFF TABLES
• The consequence resulting from a specific combination
of a decision alternative and a state of nature is a payoff.
4
School of the Built Environment
EXPECTED VALUE APPROACH
• If probabilistic information regarding the states of nature
is available, one may use the expected value (EV)
approach.
5
School of the Built Environment
EXPECTED VALUE APPROACH
• The expected value of a decision alternative is the sum
of weighted payoffs for the decision alternative.
6
School of the Built Environment
EXPECTED VALUE APPROACH
Example – lets consider the example of a housing
contractor who builds residential houses. The contractor is
planning for the number of houses to be built for the
coming year. Assume that each house costs $15,000 and
sells for $0,000 and that the probability distribution of the
market demand in this area for new houses is know as:
0 0 0 0 0
10 -150 350 350 350
20 -300 200 700 700
30 -450 50 550 1050
8
School of the Built Environment
EXPECTED VALUE APPROACH
The probability distribution of market demand is added to
the payoff table to calculate the EV of each decision
0 0 0 0 0 0
10 -150 350 350 350 250
20 -300 200 700 700 300
30 -450 50 550 1050 200
Prob DD 0.2 0.4 0.3 0.1
– Step 2:
Compute the expected value of these optimal returns.
– Step 3:
Subtract the EV of the optimal decision from the
amount determined in step (2).
11
School of the Built Environment
EXPECTED VALUE OF
PERFECT INFORMATION
Decision State of nature
(houses (Sold houses)
to be
built) 0 10 20 30 EV
0 0 0 0 0 0
10 -150 350 350 350 250
20 -300 200 700 700 300
30 -450 50 550 1050 200
Prob DD 0.2 0.4 0.3 0.1
Optimal 0 350 700 1050 455
12
School of the Built Environment
DECISION TREES
• A decision tree is a chronological representation of the
decision problem.
13
School of the Built Environment
DECISION TREES
• A decision tree is read from left to right.
14
School of the Built Environment
DECISION TREES
• The branches leaving each round node represent the
different states of nature while the branches leaving
each square node represent the different decision
alternatives.
15
School of the Built Environment
DECISION TREES
State of Payoffs
Decision node nature nodes
Decision alternatives:
1. Develop temp sensor
2. Develop pressure sensor
3. Do neither
States of nature:
1. Success
2. Failure 16
School of the Built Environment
DECISION TREES
WITHOUT PROBABILITIES
• Three commonly used criteria for decision making when
probability information regarding the likelihood of the
states of nature is unavailable are:
17
School of the Built Environment
OPTIMISTIC APPROACH
MAXIMAX
• Used by an optimistic decision maker.
18
School of the Built Environment
CONSERVATIVE APPROACH
MAXIMIN
• Used by a conservative decision maker.
21
School of the Built Environment
EXAMPLE 1
4
S1
S2 4
S3
-2
d1 0
S1
d2 S2 3
S3
-1
d3 1
S1
S2 5
S3
-3
22
School of the Built Environment
EXAMPLE 1 - MAXIMAX
• An optimistic decision maker would use the optimistic
(maximax) approach. We choose the decision that has
the largest single value in the payoff table.
Maximax
decision Maximax
payoff 23
School of the Built Environment
EXAMPLE 1 - MAXIMAX
4
S1
4 S2 4
S3
-2
d1 0
S1
d2 3
S2 3
S3
-1
d3 1
S1
5
S2 5
S3
-3
24
School of the Built Environment
EXAMPLE 1 - MAXIMIN
• A conservative decision maker would use the
conservative (maximin) approach. List the minimum
payoff for each decision. Choose the decision with the
maximum of these minimum payoffs
Minimum
Payoff
-2
-1
-3
Maximin Maximin
decision payoff
25
School of the Built Environment
EXAMPLE 1 - MAXIMIN
4
S1
-2 S2 4
S3
-2
d1 0
S1
d2 -1
S2 3
S3
-1
d3 1
S1
-3
S2 5
S3
-3
26
School of the Built Environment
EXAMPLE 1 – MINIMAX REGRET
• For the minimax regret approach, first compute a regret
table by subtracting each payoff in a column from the
largest payoff in that column.
States of nature
Decision S1 S2 S3
d1 4 4 -2
d2 0 3 -1
d3 1 5 -3
Column Max 4 5 -1
Regret table S1 S2 S3
d1 0 1 1 =-1-(-2)
=4-0
d2 4 2 0
27
d3
School of the Built Environment 3 0 2
EXAMPLE 1 – MINIMAX REGRET
• For each decision list the maximum regret. Choose the
decision with the minimum of these values.
.
Minimax
regret
decision
28
School of the Built Environment
EXAMPLE 1 - MINIMAX REGRET
4
S1
1 S2 4
S3
-2
d1 0
S1
d2 4
S2 3
S3
-1
d3 1
S1
3
S2 5
S3
-3
29
School of the Built Environment
DECISION TREES
WITH PROBABILITIES
• Decision trees are also used for displaying decision
problems with uncertainty.
30
School of the Built Environment
EXAMPLE 2
Company ABC has developed a new line of products. Top management
is attempting to decide on the appropriate marketing and production
strategy. 3 strategies are being considered: A (Aggressive), B (basic) and
C (Cautious). The market conditions under study are denoted by S
(Strong) or W (Weak). Management's best estimate of the net profits (in
$, mil) in each case is given in the following payoff table.
State of nature
Decision
S W
A 30 -8
B 20 7
C 15 10
30
P=0.45
EV=9.1
-8
P=0.55
A 20
P=0.45
B EV=12.85
7
P=0.55
C P=0.45 15
EV=12.25
P=0.55 10
32
School of the Built Environment
SENSITIVITY ANALYSIS
• Other critical points of the studied decision can be
concluded from the decision tree analysis.
33
School of the Built Environment
SENSITIVITY ANALYSIS
• Similarly solving for EVB & EVC gives:
EVC = 10 + 5P(S)
34
School of the Built Environment
SENSITIVITY ANALYSIS
EV
35
C strategy A strategy
30 B strategy
P(S) < 0.38 P(S) > 0.6
25
20
15
10
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
P(S)
-5
-10
35
School of the Built Environment
CLASS EXERCISE 1
36
School of the Built Environment
CLASS EXERCISE 1
Through an assessment of the economic environment, the
probability estimates of the likelihood of each of the 3 economic
situations under consideration is:
Given that the probability of S1, S2 and S3 is 0.4, 0.2 and 0.4
respectively, calculate the expected value for each decision & the
expected value of perfect information.
38
School of the Built Environment
CLASS EXERCISE 3
Alternative S1 S2 S3
Install 15,000 15,000 65,000
Do nothing 0 20,00 100,000
40
School of the Built Environment
CLASS EXERCISE 4
Two pumping systems “A” and ”B” are proposed for supplying
water to a residential area. The construction costs for system
“A” is $250,000 and for system “B” is $750,000. if partial failure
occurs, it is expected that the damage cost for system “A” is
$80,000 and for system “B is $15,000. If complete failure
occurs, it is expected that the damage cost for system “A” is
$450,000 and for system “B” is $400,000. the probabilities of
partial and complete failures are 5% and 1% respectively.
43
School of the Built Environment
THE END
Any questions?
44
School of the Built Environment