QBA
Decision
Analysis
PRESENTED BY:
DR SAMAR ELTANBOULY
> Introduction
>The Six Steps in Decision Making
> Types of Decision-Making Environments
Decision Making Under Uncertainty
eacaoMe ns » Decision Making Under Risk
>A Minimization Example
> Using Software for Payoff Table Problems
> Decision Trees
>How Probability Values Are Estimated by Bayesian Analysis
> Utility TheoryIntroduct
The success or failure of a person is mainly
depending on decision he or she made.
Decision making process helps people to
make their decisions & choose among
alternatives.
-
aa
'
2 t “Decision theory is an analytic and systematic
a approach to the study of decision-making
process carried out by managers to choose
among multiple alternatives.
Good decision is based on
@ Logic (consider all possible outcomes).
@ Studying all the available data & possible alternatives.
|X Using the suitable quantitative techniques.Steps of Decision-Making Process
1) Define the problem in hand
The Problem should be clearly defined and stated.
2) List the Possible Alternatives.
¥ Provided that one of the biggest mistakes that the decision makers make is to
leave out some important alternatives.
3) Identify the Possible Outcomes or States of Nature.
Yin this concem possible outcomes may be favorable or unfavorable
“Optimistic decision makers tend to ignore bad outcomes, whereas pessimistic
managers may discount a favorable outcome.
In decision theory, those outcomes over which the decision maker has little, or
no control are called states of nature
Steps of Decision-Making Process
4) List the payoff (profits) of each combination of alternatives and outcomes.
This step can be carried out through preparing the decision (payoff) table
Y Conditional values: (the payoffs results from each possible combination of
alternatives and outcomes)
5) Select one of the mathematical decision theory models.
¥ Selecting the model depends on the environment in which the organization is
operating, whether itis certain, risky, or uncertain.
6) Apply the model and make your deExample of Thompson Lumber Company
+ The Problem that John Thompson identifies is whether to expand his product
line by manufacturing and marketing a new product, backyard storage sheds.
AEE Suy
ene
Construct a large plant 180,000
Construct a small plant 20,000
Do nothing 0 0
Types of Decision-Making Environments
Decision-making under _* The consequences of every alternative are
carakcty known
+ There are several possible outcomes to
LESSTUT Tu auicg§) each decision, but the probabilities of
each outcome are known
There are several possible outcomes to
Cee ag
et ae each decision, but the probabilities of
each outcome are unknownDecision-
Making
under
Certainty
Pam Pte ela maton -me lacie
certainty
“ Decision maker is certain about the
consequence of every alternative.
+ Let's say you have $1000 to invest for a year
period:
1) To open saving account paying 6% interest
2) To invest in a treasury bond paying 10%
interest
“Both investment are safe and secured
“ Solution: The treasury bond will pay a higher
return of $100 better than $60.Decision-
Making
under Risk
“Decision theory models for business
problems in this environment typically
employ 2 equivalent criteria
1) Maximization of expected monetary
value.
2)Minimization of expected opportunity
loss.
+ Decision making under risk is a
probabil
ic decision situation. Several
possible states of nature may occur,
each with a given probability.
* Decision maker will select the
alternative with the highest expected
monetary value and lowest expected
opportunity losses.A. Expected Monetary Value (EMV):
v Itis defined as the weighted sum of
possible payoffs for each alternative.
Y Expected value (Mean Value): is the
long-run average value of that decision
v EMV for an alternative is just the sum of
possible payoffs of the alternative, each
weighted by the probability of that
payoff occurring
A. Expected Monetary Value (EMV):
EMV (alternative ¢) = Y X;P(X;)
X; = payoff for the alternative in state of nature 4
P(X,) = probability of achieving payoff X; (state of nature <)
= Summation symbol
EMV 1alternative2
= (Payoff in 1st state of nature) X (Probability of 1st state of nature)
+ (Payoff in 2nd state of nature) X (Probability of 2nd state of nature)
+... + (Payoff in last state of nature) X (Probability of last state of nature)|(3200, 000)(0.: 50) + +
ENTE) [= \ a) POC) A (-$180,000)(0.50) = $40,600
(3) ($) > EMV (small plant) =
large plant 200,000 -180,000 [syasuile soy
Small plant sooeee -20,000 > EMV (do nothing) =
DONT 6 ($0)(0.50) + ($0)(0.50) = 86
B. Expected Value with Perfect Information (EVwPI):
v Itis the expected or average return, in the long run, if we have
perfect information before a decision should be made.
EV wPI =5 (Best payoff in state of nature i) (probability of state of nature i)
+ BV wPI
Rest payoff in first state of nature) x (Probability of first state of nature)
+(Best payoff in Second state of nature) x (Probability of second state of nature)
+ + (Best payoff in last state of nature) x (Probability of last state of nature)
Inthe Example: EVwPI = ($200, 000)(0.5) + ($0)(0.5) = $100,000C. Expected Value of Perfect Information (EVPI):
Y tis the expected value with perfect information minus the expected
value without perfect information (the best or maximum EMV).
v_EVPI places an upper ceiling on what to pay in return of information
EVPI = EVwPI — Best EMV
In the Example: EVP] = EVwPI — Maximum EMV
} [Olt UnUee)
Aenea
eens
ied
(3)
Large plant 200,000 -180,000 10,000
Smallplant 100,000 -—-20,000 40,000
Do nothing 0 0 0
Probabilities. 50% 50%D. Expected Opportunity Loss (EOL) (Regret):
v Itis an alternative approach to minimize
expected opportunity loss (EOL)
Y EOLis the cost of not picking the best solution.
¥ The minimum (EOL) can be identified through
creating the opportunity loss table.
Ist Step: To subtract each pay off in the column
(under each state of nature) from the best
payoff in the same column.
$ _ Incaseof favorable market, we subtract the payotfs of each alternative from the best payoff ($200,000).
¥ _ In case of unfavorable market, we subtract the payoffs of each alternative from the best pay off ($0).
Be
15* Step to
calculate ari
EOL Large plant 200,000-200,000= 0~(-180,000)
° 180,000
‘Small plant 200,000- 100,000= ‘0-(-20,000)=
100,000 20,000
Do nothing 200,000- 0 = oa
200,000 Cmte
Probabilities 50% 50%2°4 Step to
calculate EOL EMV (large plant) = (0)(0.50) +
($180,000}(0.50) =
Bae
Favorable | Unfavorable
Market ry
ean iS)
EMV (small plant) =
($100,000)(0.50) +
EeWO Me (520,000)(0.50) = $60,000
Large plant 0 180,000 90,000 Emry (do nothing) =
‘Small plant 100,000 20,000 60,000 a ,000}(0.50) + ($0)(0.50) =
Do nothing 200,000 0 100,000
Probabilities 50% 50% EMV =$60,000 is the lowest
EVPI = Minimum EO.
Example (2)
Maria Rojas is considering the possibility of opening a small dress shop on
Fairbanks Avenue, a few blocks from the university. She has located a
good mall that attracts students.
Her options are to open a small shop, a medium-sized shop, or no shop at
all. The market for a dress shop can be good, average, or bad.
The probabilities for these three possibilities are 0.2 for a good market,
0.5 for an average market, and 0.3 for a bad market.
The net profit or loss for the medium-sized and small shops for the various
market conditions are given in the following table. Building no shop at all
yields no loss and no gain.Ua ECs Pond
ey -40000
Wee) 60000
Per) °
eet . . 03
a) What do you recommend?
b) Calculate the EVPI.
) Develop the opportunity loss table for this situation, What decisions would be made
using the minimum EOL criterion?
[EEE
Solution
Lilet em eI CT MMe ote Meade ay
* EMV (1) = (75000*0.2) + (25000*0.5) + (-40000*0.3)=
* EMV (2) = (100000*0.2) + (35000*0.5) + (-60000*0.3) =$ 19500
* EMV (N) = (0*0.2) + (0*0.5) + (0*0.3) =$0
Based on EMV, building a medium sized shop will be
Te a alaSecond: Calculating the opportunity loss table.
Alternative! Good | Average | Bad Market|
IVE IE
Bone!
ach
GSE} 25000 10000 40000
Pon eacons
coco
0 0 60000
00*0.2) +
(o*03)
100000 35000 0 Based on EOL, building a medium
sized shop will be the best choice.
Seren
Calculati i meme value of fi information
EVwPI = (100000*0.2) + (35000* 0.5) + (0*0.3)
VEZ cet )
NCR ne eeeSensitivity
Analysis
Sensitivity Analysis
> Sensitive analysis investigates how our decision
might change with different input data.
P = probability of a favorable market
> EMV (large plant) = $200,000 P- $180,000 (1-P) =
$380,000 P- $180,000
EMV (small plant) = $100,000 P- $20,000 (1-P) =
$120,000 P- $20,000
EMV (Do nothing) = $0 P- $0 (1-P) = $0oe than 0.167
ena 0167-0615
Construct large plant Greater than 0.615
Decision-
Making
under
UncertaintyDecision Making Under Uncertainty
Several criteria exist for making decisions
Peet eae ecm
Rieter Rees
‘L.Optimistic (Maximax)
2.Pessimistic (Maximin)
3.Criterion of realism (Hurwicz)
4.Equally likely (Laplace)
5.Minimax regret
sO Re) Lelia
Maximax)
. Ameen
le Cele
sll Market | Market | maximax
nea rear 5) (Ce ned)
We choose, the best
(maximum) payoff for Large plant $200,000 180,000
each alternative is Small lant 100,000 -20,000 100,000
considered, and the
alternative with the Bonothing a g o
best (maximum) of
these is selected.
‘Maximax is an2) Pessimistic
(Maximin)
worst (minimum)
payoff for each
alternative is
considered, and the
alternative with the
best (maximum) of
these is selected
3) Criterion of Realism
(Hurwicz Criterion)
[/ Criterion of realism uses the|
‘weighted average approach.
JV itis a compromise between
fan optimistic and a
pessimistic decision.
J’ The advantage of this
approach is that it allows
the decision maker to build
in personal feelings about
relative optimism and
pessimism,
Ameen
Coca cy
Cram meLCame VTC
o) (of) CE)
Large plant $200,000 -180,000 ~180,000
‘Small plant 100,000 -20,000 -20,000
Do nothing 0 0 0
weighted average ~ a(Best in row) + (1 ~ a)(Worst in row)
Bre ee
ead
Ne aes bid
(o) (o} 3
ESE 0.8) ()
Large plant $200,000 -180,000
‘Small plant 100,000 -20,000
eer
Do nothing 0 ° 0OE el) VAM Col
(Laplace)
[” Equally likely criterion uses
the average outcome.
JV This involves finding the
average payoff for each
alternative and selecting
the alternative with the
best or highest average.
JV ttassumes that all
probabilities of occurrence|
for the states of nature are
equal, and thus each state
of nature is equally ikely
CyB cbantetcia
(EOL)
JY Minimax regret criterion is
based on opportunity loss.
IV Opportunity loss refers to
the difference between the}
optimal profit or payoff for
a given state of nature and
the actual payoff received
for a particular decision for
that state of nature.
|’ Itis the amount lost by not
picking the best alternative}
ina given state of nature
Large plant
‘Small plant
Do nothing
Large plant
‘Small plant
Do nothing
Been
ee
WES Oc
(3) 3)
eT
LMT tcd
(s)
$200,000 —-180,000
100,000 20,000 40,000
o 0 0
Been
Ec eM Merc os
Dec
ny Hy IIS
TO}
100,000 20,000
200,000 0 200,000Decision Tree
Decision Trees
1) Define the problem.
2), Structure or draw the decision tree.
3) Assign probabilities to the states of nature.
4) Estimate payoffs for each possible combination
of alternatives and states of nature
5) Solve the problem by computing EMVs for each
state-of-nature node. This is done by working
backward, that is, starting at the right of the
tree and working back to decision nodes on the
left. Also, at each decision node, the alternative
with the best EMV is selected.a larolal essen
Decision Tree
A decision node: from
crs
eee
Pee eee
Can
occur.
Example on
Decision Tree
Thompson’s Decision Tree
A Siate-of.Nature Node
Favorable Market
ADecislonNode
Ee
Snel Fane
Thompson’s Decision Tree
+= (0.5)200,000) + (0.5)(-$180,000)
Payotts
$200,000
$20,000
(0.5($100,000)
+ (0:5)(-$20,000)
#Example on
Thompson’s
1D-el (eo) ga =12
Seo
on
Pore eens
eet
1. Given favorable survey results,
¥ EMV (node 2) = EMV (large plant|positive survey)
= (0.75)($190,000) + (0.25)(—$190,000)
= $95,000
¥ EMV (node 3) = EMV (smail plant|positive survey)
= (0.75)($90,000) + (0.25)(—$30,000) = $60,000
2. Given negative survey results,
v EMV (node 4) = EMV (large plant|negative survey)
= (0.25)($190,000) + (0.75)(—$190,000) = -$95,000
~ EMV (node 5) = EMV(small plant\negative survey)
= (0.25)($90,000) + (0.75)(—$30,000) = $03. Continuing the upper part of the tree and moving backward, we compute the
expected value of conducting the market survey.
EMV (node 1) = EMV (conduct survey)
(0.5)($95,000) + (0.5)($0) = $95,000
4, Ifthe market survey is not conducted,
EMV (node 6) = EMV large plant)=$10,000
¥ EMV (node 7) = EMV (smail plant)=$40,000
¥ EMV of no plant is$O
5. Expected value of Sample information (EVSI)
EVSI = (EV with SI + cost) — (EV without ST)
EYVSI = ($47,500 + $10,000) — $40,000 = $17,500
. VSI 17,500 =
6. Efficiency of Sample Information = "100% = =" 100% = 29%
The market survey is 108% as efficient as perfect information