Professional Documents
Culture Documents
DECISION ANALYSIS
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
Table of Contents
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
Learning Objectives
5. Apply Bayes’ decision rule to solve a decision analysis problem.
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
1. Decision Analysis
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
1. Decision Analysis
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
1. Decision Analysis
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
1. Decision Analysis
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
1. Decision Analysis
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
2. The Goferbroke Company Problem
The Goferbroke Company develops oil wells in
unproven territory.
A consulting geologist has reported that there is a
one-in-four chance of oil on a particular tract of
land.
Drilling for oil on this tract would require an
investment of about $100,000.
If the tract contains oil, it is estimated that the net
revenue generated would be approximately
$800,000.
Another oil company has offered to purchase the
tract of land for $90,000.
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
2.1. Prospective Profits
Profit
Alternative
Chance of status 1 in 4 3 in 4
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
2.2. Decision Analysis Terminology
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
2.2. Decision Analysis Terminology
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
2.3. Prior Probabilities
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
2.4. Payoff Table (Profit in $Thousands)
State of Nature
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
3. Decision Criteria
Maximax Maximin
Criterion Criterion
Maximum Bayes’
Likelihood Decision
Criterion Rule
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
3.1. The Maximax Criterion
The maximax criterion is the decision criterion
for the eternal optimist.
It focuses only on the best that can happen.
Procedure:
Identify the maximum payoff from any state of
nature for each alternative.
Find the maximum of these maximum payoffs
and choose this alternative.
State of Nature
Alternative Oil Dry Maximum in Row
Drill for oil 700 –100 700 Maximax
Sell the land 90 90 90
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
3.2. The Maximin Criterion
The maximin criterion is the decision criterion
for the total pessimist.
It focuses only on the worst that can happen.
Procedure:
Identify the minimum payoff from any state of
nature for each alternative.
Find the maximum of these minimum payoffs
and choose this alternative.
State of Nature
Alternative Oil Dry Minimum in Row
Drill for oil 700 –100 –100
Sell the land 90 90 90 Maximin
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
3.3. The Maximum Likelihood Criterion
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
3.4. Bayes’ Decision Rule
A B C D E F
1 Bayes' Decision Rule for the Goferbroke Co.
2
3 Payoff Table State of Nature Expected
4 Alternative Oil Dry Payoff
5 Drill 700 -100 100
6 Sell 90 90 90
7
8 Prior Probability 0.25 0.75
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
3.4. Bayes’ Decision Rule
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
3.4. Bayes’ Decision Rule
700
Oil (0.25)
Drill
Dry (0.75)
-100
A
Sell
90
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
4. Using RSPE to Construct Decision Trees
A B C D E F G
1
2 Sell
3 90
4 90 90
5 1
6 90
7 Decision 2
8 0
9 0 0
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
4. Using RSPE to Construct Decision Trees
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
4. RSPE Results
The numbers inside each decision node indicate
which branch should be chosen (assuming the
branches are numbered consecutively from top to
bottom).
The numbers to the right of each terminal node is
the payoff if that node is reached.
The number 100 in cells A10 and E6 is the
expected payoff at those stages in the process.
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
4. RSPE Results
A B C D E F G H I J K
1 25%
2 Oil
3 700
4 Drill 800 700
5
6 -100 100 75%
7 Dry
8 -100
9 1 0 -100
10 100
11
12 Sell
13 90
14 90 90
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
4. Consolidate the Data and Results
A B C D E F G H I J K
1 25%
2 Oil
3 700
4 Drill 800 700
5
6 -100 100 75%
7 Dry
8 -100
9 1 0 -100
10 100
11
12 Sell
13 90
14 90 90
15
16
17 Data
18 Cost of Drilling 100
19 Revenue if Oil 800
20 Revenue if Sell 90
21 Revenue if Dry 0
22 Probability of Oil 25%
23
24 Action Drill
25
26 Expected Payoff 100
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
5. Sensitivity Analysis: Prior Probability of Oil = 0.15
A B C D E F G H I J K
1 15%
2 Oil
3 700
4 Drill 800 700
5
6 -100 20 85%
7 Dry
8 -100
9 2 0 -100
10 90
11
12 Sell
13 90
14 90 90
15
16
17 Data
18 Cost of Drilling 100
19 Revenue if Oil 800
20 Revenue if Sell 90
21 Revenue if Dry 0
22 Probability of Oil 15%
23
24 Action Sell
25
26 Expected Payoff 90
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
5. Sensitivity Analysis: Prior Probability of Oil = 0.35
A B C D E F G H I J K
1 15%
2 Oil
3 700
4 Drill 800 700
5
6 -100 20 85%
7 Dry
8 -100
9 2 0 -100
10 90
11
12 Sell
13 90
14 90 90
15
16
17 Data
18 Cost of Drilling 100
19 Revenue if Oil 800
20 Revenue if Sell 90
21 Revenue if Dry 0
22 Probability of Oil 15%
23
24 Action Sell
25
26 Expected Payoff 90
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
5. Using Data Tables to Do Sensitivity Analysis
A B C D E F G H I J K L M
1 15%
2 Oil
3 700
4 Drill 800 700
5
6 -100 20 85%
7 Dry
8 -100
9 2 0 -100
10 90
11
12 Sell
13 90
14 90 90
15
16 Probability Expected
17 Data of Oil Action Payoff
Select these cells before
18 Cost of Drilling 100 Sell 90
choosing Data Table
19 Revenue if Oil 800 15%
from the What-If
20 Revenue if Sell 90 17% Analysis menu on the
21 Revenue if Dry 0 19% Data tab.
22 Probability of Oil 15% 21%
23 23%
24 Action Sell 25%
25 27%
26 Expected Payoff 90 29%
27 31%
28 33%
29 35%
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
5. Data Table Results
The Effect of Changing the Prior Probability of Oil
I J K
16 Probability Expected
17 of Oil Action Payoff
18 Drill 100
19 0.15 Sell 90
20 0.17 Sell 90
21 0.19 Sell 90
22 0.21 Sell 90
23 0.23 Sell 90
24 0.25 Drill 100
25 0.27 Drill 116
26 0.29 Drill 132
27 0.31 Drill 148
28 0.33 Drill 164
29 0.35 Drill 180
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
6. Checking Whether to Obtain More Information
Might it be worthwhile to spend money for more information
to obtain better estimates?
A quick way to check is to pretend that it is possible to
actually determine the true state of nature (“perfect
information”).
EP (with perfect information) = Expected payoff if the
decision could be made after learning the true state of nature.
EP (without perfect information) = Expected payoff from
applying Bayes’ decision rule with the original prior
probabilities.
The expected value of perfect information is then
EVPI = EP (with perfect information) – EP (without
perfect information).
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
6. Expected Payoff with Perfect Information
A B C D E F G H I J K
1
2 Drill
3 25% 700
4 Oil 700 700
5 1
6 0% 700
7 Sell
8 90
9 90 90
10
11 242.5
12 Drill
13 75% -100
14 Dry -100 -100
15 2
16 0 90
17 Sell
18 90
19 90 90
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
7. Using New Information to Update the Probabilities
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
7.1. Seismic Survey for Goferbroke
P(finding | state)
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
7.2. Calculating Joint Probabilities
Each combination of a state of nature and a
finding will have a joint probability determined by
the following formula:
P(state and finding) = P(state) P(finding | state)
P(Oil and FSS) = P(Oil) P(FSS | Oil) = (0.25)(0.6)
= 0.15.
P(Oil and USS) = P(Oil) P(USS | Oil) = (0.25)
(0.4) = 0.1.
P(Dry and FSS) = P(Dry) P(FSS | Dry) = (0.75)
(0.2) = 0.15.
P(Dry and USS) = P(Dry) P(USS | Dry) = (0.75)
(0.8) = 0.6.
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
7.3. Probabilities of Each Finding
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
7.4. Calculating the Posterior Probabilities
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
7.5. Probability Tree Diagram
Prior Conditional Joint Posterior
Probabilities Probabilities Probabilities Probabilities
P(state) P(finding | state) P(state and finding) P(state | finding)
0.15 = 0.5
0.25(0.6) = 0.15 0.3
0.6 Oil and FSS Oil, given FSS
FSS, given Oil
0.4
0.25 USS, given Oil 0.1 = 0.14
0.25(0.4) = 0.1 0.7
Oil
Oil and USS Oil, given USS
0.15 = 0.5
0.2 0.75(0.2) = 0.15 0.3
0.75 Dry, given FSS
Dry FSS, given Dry Dry and FSS
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
7.6. Posterior Probabilities
P(state | finding)
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
7.7. Template for Posterior Probabilities
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
7.7. Template for Posterior Probabilities
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
8. Decision Tree for the Full Goferbroke Co.
Problem
Oil
f
Drill
Dry
c
Unfavorable Sell
Oil
b g
Drill Dry
Do seismic survey
Favorable
d
a Sell
Oil
h
No seismic survey Drill
Dry
e
Sell
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
8.1. Decision Tree with Probabilities and
Payoffs
Payoff
Oil (0.143) 670
800
f
Drill 0
-100 Dry(0.857) -130
c
90
Unfavorable Sell 60
0
Oil (0.5) 670
b 800
g
Do seismic survey 0 Drill 0
Favorable -100 Dry (0.5) -130
-30 (0.3)
d
90
a Sell 60
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
8.2. The Final Decision Tree
Payoff
Oil (0.143) 670
-15.7
800
f
0
Drill
-100 Dry (0.857) -130
60
c
90
Unfavorable Sell 60
0
123 Oil (0.5) 670
270
b g 800
0 Drill 0
Do seismic survey -100 Dry (0.5) -130
-30 Favorable (0.3) 270
d
123 90
a Sell 60
Oil (0.25) 700
100
0 800
h
No seismic survey Drill 0
-100 Dry (0.75) -100
100
e
90
Sell 90
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
8.3. RSPE for the Full Goferbroke Co.
Problem 14%
Oil
670
Drill 800 670
Sell
60
90 60
Do Survey
50%
-30 123 Oil
670
Drill 800 670
1 Sell
123 60
90 60
25%
Oil
700
Drill 800 700
Sell
90
90 90
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
9. Organizing the Spreadsheet for Sensitivity
Analysis
Decision Tree for Goferbroke (with Survey) Prior
Probability Do If Survey
14% Data of Oil Survey? Favorable
Oil Cost of Survey 30 Yes Drill
670 Cost of Drilling 100 0 No Sell
Drill 800 670 Revenue if Oil 800 0.1 No Drill
Revenue if Sell 90 0.2 Yes Drill
-100 -15.71428571 86% Revenue if Dry 0 0.3 Yes Drill
70% Dry Probability of Oil 0.25 0.4 No Drill
Unfavorable -130 P(FSS | Oil) 0.6 0.5 No Drill
2 0 -130 P(USS | Dry) 0.8 0.6 No Drill
0 60 0.7 No Drill
0.8 No Drill
Sell Action 0.9 No Drill
60 Do Survey? Yes 1 No Drill
90 60
Do Survey If No If Yes
50%
-30 123 Oil Drill Drill If Favorable
670 Sell if Unfavorable
Drill 800 670
1 Sell
123 60 Data: P(Finding | State)
90 60 State of Prior Finding
Nature Probability FSS USS
25% Oil 0.25 0.6 0.4
Oil Dry 0.75 0.2 0.8
700
Drill 800 700
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
HƯỚNG DẪN
1. NHẬP DATA
2. KHAI BÁO DATA TRÊN TEMPLATE
3. KHAI BÁO DATA TRÊN CÂY TỪ DATA
TABLE VÀ TEMPLATE
4. LẬP BẢNG PRIOR % VÀ KHAI CÁC
QUYẾT ĐỊNH THEO CÂY
5.CHỌN BẢNG VÀ VÀO DATA TABLE
TRÊN MENU WHAT-IF ANALYSIS TRÊN
DATA, CHỌN THAY ĐỔI COLUMN...BẤM
OK
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
9. Data Table: Optimal Policy vs. Prior
Probability of Oil
Y Z AA AB AC AD
1 Prior Expected
2 Probability Do If Survey If Survey If No Payoff
3 of Oil Survey? Favorable Unfavorable Survey ($thousands)
4 Yes Drill Sell Drill 123
5 0 No Sell Sell Sell 90
6 0.1 No Drill Sell Sell 90
7 0.2 Yes Drill Sell Sell 102.8
8 0.3 Yes Drill Sell Drill 143.2
9 0.4 No Drill Drill Drill 220
10 0.5 No Drill Drill Drill 300
11 0.6 No Drill Drill Drill 380
12 0.7 No Drill Drill Drill 460
13 0.8 No Drill Drill Drill 540
14 0.9 No Drill Drill Drill 620
15 1 No Drill Drill Drill 700
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
9. Optimal Policy
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10. Using Utilities to Better Reflect the Values of
Payoffs
Thus far, when applying Bayes’ decision rule, we
have assumed that the expected payoff in
monetary terms is the appropriate measure.
In many situations, this is inappropriate.
Suppose an individual is offered the following
choice:
Accept a 50-50 chance of winning $100,000.
Receive $40,000 with certainty.
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10. Using Utilities to Better Reflect the Values of
Payoffs
Many would pick $40,000, even though the
expected payoff on the 50-50 chance of winning
$100,000 is $50,000. This is because of risk
aversion.
A utility function for money is a way of
transforming monetary values to an appropriate
scale that reflects a decision maker’s preferences
(e.g., aversion to risk).
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.1. A Typical Utility Function for Money
U(M)
0.75
0.5
0.25
0
$10,000 $30,000 $60,000 $100,000 M
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.2. Shape of Utility Functions
M M M
(a) Risk averse (b) Risk seeker (c) Risk neutral
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.3. Utility Functions
When a utility function for money is incorporated into a
decision analysis approach, it must be constructed to fit
the current preferences and values of the decision maker.
Fundamental Property: Under the assumptions of
utility theory, the decision maker’s utility function for
money has the property that the decision maker is
indifferent between two alternatives if the two
alternatives have the same expected utility.
When the decision maker’s utility function for money is
used, Bayes’ decision rule replaces monetary payoffs by
the corresponding utilities.
The optimal decision (or series of decisions) is the one
that maximizes the expected utility.
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.4. Illustration of Fundamental Property
By the fundamental property, a decision maker with the utility
function below-right will be indifferent between each of the three
pairs of alternatives below-left.
U(M)
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.5. The Equivalent Lottery Method
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.5. The Equivalent Lottery Method
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.6. Generating the Utility Function for Max
Flyer
The possible monetary payoffs in the Goferbroke
Co. problem are –130, –100, 0, 60, 90, 670, and
700 (all in $thousands).
Set U(Maximum) = U(700) = 1.
Set U(Minimum) = U(–130) = 0.
To find U(M), use the equivalent lottery method.
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.6. Generating the Utility Function for Max
Flyer
For example, for M=90, consider the two
alternatives:
A1: Obtain a payoff of 700 with
probability p
Obtain a payoff of –130 with probability
1–p.
A2: Definitely obtain a payoff of 90
If Max chooses a point of indifference of p = 1/3,
then U(90) = 1/3.
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.7. Max’s Utility Function for Money
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.8. Utilities for the Goferbroke Co. Problem
–130 0.00
–100 0.05
60 0.30
90 0.33
670 0.97
700 1.00
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.9. Decision Tree with Utilities
14%
Oil
0.97
Drill 0.97 0.97
0 0.138571429 86%
70% Dry
Unfavorable 0
2 0 0
0 0.3
Sell
0.3
0.3 0.3
Do Survey
50%
0 0.3555 Oil
0.97
Drill 0.97 0.97
0 0.485 50%
30% Dry
Favorable 0
1 0 0
0 0.485
1 Sell
0.3555 0.3
0.3 0.3
25%
Oil
1
Drill 1 1
0 0.2875 75%
Dry
No Survey 0.05
2 0.05 0.05
0 0.333
Sell
0.333
0.333 0.333
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.10. Exponential Utility Function
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.10. Exponential Utility Function
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.11. Setting up RSPE to use the Exponential
Utility Function
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
10.11. Decision Tree with an Exponential
Utility Function 14%
Oil
670
Drill 800 670
0.488291422
-100 -48.06593977 86%
70% -0.04923984 Dry
Unfavorable -130
2 0 -130
0 60 -0.138828383
0.058235466
Sell
60
90 60
Do Survey 0.058235466
50%
-30 97.81602738 Oil
0.093184282 670
Drill 800 670
0.488291422
-100 192.0465146 50%
30% 0.174731519 Dry
Favorable -130
1 0 -130
0 192.0465146 -0.138828383
0.174731519
1 Sell
97.81603 60
0.093184 90 60
0.058235466
25%
Oil
700
Drill 800 700
0.503414696
-100 48.11465215 75%
0.046975485 Dry
No Survey -100
2 0 -100
0 90 -0.105170918
0.086068815
Sell
90
90 90
0.086068815
Risk Tolerance 1000
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013
Q&A
UEH – School of International Business & Marketing © The McGraw-Hill Companies, Inc., 2013