Lecture 2 & 3 - Chapter 3 - Decision Analysis
Lecture 2 & 3 - Chapter 3 - Decision Analysis
Meaningful information
DECISION ANALYSIS
CERTAINTY
UNCERTAINTY
RISK
Type of Decision Making Environment
Ex: $1000 invest for 1-year 1. Maximax (optimistic) 1. Expected monetary value
period 2. Maximin (pessimistic) (EMV)
Alternatives: invest in gov bond: 3. Criterion of realism 2. Expected value of perfect
5%/year, open a saving account: (Hurwicz) information (EVPI)
6%/year 4. Equally likely (Laplace) 3. Expected value with
5. Minimax Regret perfect information
(EVwPI)
4. Expected opportunity loss
Six steps in decision a
3. Identify
1. Define 6. Apply
2. List the possible
the 4. List the 5. Select a the model
possible outcomes
problem/ payoffs model to make
alternatives or state of
obj decision
nature
Thompson Lumber
company STATE OF NATURE
FAVORABLE UNFAVORABLE
ALTERNATIVE MARKET ($) MARKET ($)
Do nothing 0 0
Decision making under uncertainty
STATE OF NATURE
FAVORABLE UNFAVORABLE ROW
ALTERNATIVE MARKET ($) MARKET ($) AVERAGE ($)
Construct a large –
plant 200,000 180,000 10,000
Construct a small
–20,000
plant 100,000 40,000
Equally likely
Do nothing 0 0 0
Decision making under uncertainty
STATE OF NATURE
FAVORABLE UNFAVORABLE MAXIMUM IN
ALTERNATIVE MARKET ($) MARKET ($) A ROW ($)
Construct a large
0 180,000
plant 180,000
Construct a small
20,000
plant 100,000 100,000
Minimax
Do nothing 0
200,000 200,000
Decision making under risk
• Decision making when there are several possible states of
nature and we know the probabilities associated with each
possible state.
• Most popular method is to choose the alternative with the
highest expected monetary value (EMV)
• The expected value or the mean value (EMV) is the long run
average value of that decision.
• EMV for an alternative is just the sum of possible payoffs of
the alternative, each weighted by the probability of that
payoff occurring.
rnative i) = (payoff of first state of nature)
x (probability of first state of nature)
+ (payoff of second state of nature)
x (probability of second state of nature)
+ … + (payoff of last state of nature)
x (probability of last state of nature)
Decision making under risk
EMV for Thompson Lumber
Each market has a probability of 0.50
Which alternative would give the highest EMV?
The calculations are
EMV (large plant) = (0.50)($200,000) + (0.50)(–$180,000)
= $10,000
EMV (small plant) = (0.50)($100,000) + (0.50)(–$20,000)
= $40,000
EMV (do nothing) = (0.50)($0) + (0.50)($0) = $0
STATE OF NATURE
FAVORABLE UNFAVORABLE
ALTERNATIVE MARKET ($) MARKET ($) EMV ($)
Construct a large –
plant 200,000 180,000 10,000
Construct a small
–20,000
plant 100,000 40,000
Do nothing 0 0 0
Probabilities 0.50 0.50
Largest EMV
Decision making under risk
Expected value of perfect information
Two steps processes:
1. Determine the Expected Value with Perfect
Information
2. Compute Expected Value of Perfect Information
■ EVwPI is the long run average return if we have perfect
(EVPI)
information before a decision is made.
EVwPI = (best payoff for first state of nature)
x (probability of first state of nature)
+ (best payoff for second state of nature)
x (probability of second state of nature)
+ … + (best payoff for last state of nature)
x (probability of last state of nature)
■ EVPI places an upper bound on what you should pay for
additional information.
EVPI = EVwPI – Maximum EMV
Decision making under risk
Expected value of perfect information
■ Scientific Marketing, Inc. offers analysis that will
provide certainty about market conditions (favorable)
■ Additional information will cost $65,000
■ Is it worth purchasing the information?
With perfect information
Best alternative (max EMV) for favorable state
Thompson Lumber of nature is build a large plant with a payoff of
company STATE OF NATURE $200,000
Best alternative (max EMV) for unfavorable
UNFAVORA state of nature is to do nothing with a payoff of
FAVORABLE BLE $0
ALTERNATIVE MARKET ($) MARKET ($) EVwPI = ($200,000)(0.50) + ($0)(0.50) =
$100,000
–
Construct a large plant
200,000 180,000 The maximum EMV without additional
information is $40,000
– EVPI = EVwPI – Maximum EMV without PI
Construct a small plant
100,000 20,000
= $100,000 - $40,000
Do nothing 0 0 = $60,000
Do nothing 0
200,000 100,000
Probabilities 0.50 0.50 Minimum EOL
rge plant) = (0.50)($0) + (0.50)($180,000)
= $90,000
mall plant) = (0.50)($100,000) + (0.50)($20,000)
= $60,000
o nothing) = (0.50)($200,000) + (0.50)($0)
= $100,000
SENSITIVITY ANALYSIS
Sensitivity analysis examines how our decision might
change with different input data
For the Thompson Lumber example
–$200,000
Point 1: Point 2:
EMV(small plant) = EMV(large plant)
EMV(do nothing) = EMV(small
plant)
20,000
$120,000 P $20,000 $380,000 P $180,000
0 $120,000 P $20,000 P 0.167 160,000
120,000 P 0.615
260,000
Using Excel QM
HOMEWORK
• Assignments (Chapter 3): 3-17, 3-20, 3-23, 3-25, 3-28
and 3-29. (It is better if you can do all problems in the
textbook ^_^)
• One-page summary of Decision Analysis – Part 1
PAR
T2
DECISION TREE
1. Develop accurate and useful decision trees
2. Revise probabilities using Bayesian analysis
3. Understand the importance and use of utility theory in
decision making
4. Use computers to solve basic decision-making problems
Structure of Decision
A 3-min instruction on how to
draw a decision tree
Trees
■
https://youtu.be/ydvnVw80I_8
Trees start from left to right
■ Represent decisions and outcomes in sequential
order
■ Squares represent decision nodes
■ Circles represent states of nature nodes
■ Lines or branches connect the decisions nodes
and the states of nature
decision nodes
STATE OF NATURE
FAVORABLE UNFAVORABLE
ALTERNATIVE MARKET ($) MARKET ($)
Construct a large plant 200,000 –180,000
Construct a small plant 100,000 –20,000
Do nothing 0 0
STATE OF NATURE
FAVORABLE UNFAVORABLE
EXAMPLE 1.1 ALTERNATIVE
Construct a
MARKET ($) MARKET ($)
–180,000
large plant 200,000
Construct a
–20,000
small plant 100,000
Do nothing 0
0
$0
EXAMPLE 1.2
A MORE COMPLEX DECISION FOR
THOMPSOM LUMBER – SAMPLE
INFORMATION
Let’s say that John Thompson has two decisions to make, with the
second decision dependent on the outcome of the first. Before
deciding about building a new plant, John has the option of
conducting his own marketing research survey, at a cost at
$10,000. The information from his survey could help him decide
whether to construct a large plant, a small plant, or not to build at
all. John recognizes that such a market survey will not provide him
perfect information, but it may help quite a bit nevertheless.
• There is a 45% chance that the survey will indicate a favorable market.
• 78% is the probability of a actual favorable market given a favorable result
from the market survey.
• There is a 27% chance the market will be actually favorable given that John’s
survey results are negative.
Thompson’s Complex Decision
Tree
Plant –$30,000
M
ct
du
No Plant
–$10,000
n
Co
$106,400
a rge $63,600 Favorable Market (0.78)
L $90,000
) Small
(0
. 45 Plant 3 Unfavorable Market (0.22)
–$30,000
e y ts le
u rv sul ab No Plant
–$10,000
S Re vor
S a
1 urveF –$87,400 Favorable Market (0.27)
$190,000
y
y
ve
R
Ne esu (0.5 Pla
n t 4 Unfavorable Market (0.73)
ur
–$190,000
5) rge
tS
ga lts $2,400
$2,400
La Favorable Market (0.27)
ti v Small $90,000
ke
Plant –$30,000
M
ct
No Plant
du
–$10,000
on
$49,200
C
t g
Sur
ve y Lar $40,000 Favorable Market (0.50)
$100,000
Small
Plant 7 Unfavorable Market (0.50)
–$20,000
No Plant
$0
When EMV of conducting the market survey and not conducting the
market survey are indifferent:
If EMV (node 1) = 40
Then 104p + 2.4 = 40 and then p = 0.36
If EMV (node 1) > 40 then p >0.36 Should conduct the survey
EXAMPLE 1.3 Calculating Revised Proba
• when there was actually a favorable market for the product, 70% of the survey
correctly predicted positive results.
• On the other hand, when there was actually an unfavorable market for the
product, 80% of the survey correctly predicted negative results.
• Recall that without any market survey information, John’s best estimates of a
favorable and unfavorable market are: P(FM)= 0.5, P(UM=0.5)
STATE OF NATURE
RESULT OF FAVORABLE MARKET UNFAVORABLE MARKET
SURVEY (FM) (UM)
Positive (predicts P (survey positive | P (survey positive |
favorable market FM) UM)
for product) = 0.70 = 0.20
Negative (predicts P (survey negative P (survey negative
unfavorable | FM) | UM)
market for
product) = 0.30 = 0.80
Calculating Revised Probabilities using Bayes
theorem
(1) Or (2)
For this example,
A and A’ will represent a favorable market (FM) and an unfavorable market (UM), respectively.
B and B’ will represent a positive survey and a negative survey, respectively.
POSTERIOR PROBABILITY
CONDITIONAL
PROBABILITY P(STATE OF
STATE OF P(SURVEY POSITIVE | PRIOR JOINT NATURE | SURVEY
NATURE STATE OF NATURE) PROBABILITY PROBABILITY POSITIVE)
FM 0.70 X 0.50 = 0.35 0.35/0.45 = 0.78
UM 0.20 X 0.50 = 0.10 0.10/0.45 = 0.22
P(survey results positive) = 0.45 1.00
P ( survey positive | FM ) P ( FM )
P (FM | survey positive)
P(survey positive |FM) P(FM) P(survey positive |UM) P(UM)
Homework:
1. Assignments (Chapter 3): 3-34, 3-36, 3-37, 3-39, 3-41,
3-42, 3-48 and 3-52. (It is better if you can do all
problems in the textbook ^_^)
2. One-page summary of chapter 3 - part 2.