Académique Documents
Professionnel Documents
Culture Documents
DBM
Quantitative Techniques in Management
Semester - 1
Session - 7
STATISTICAL DECISION
THEORY
Introduction to Decision
Analysis
The field of decision analysis provides a
framework for making important decisions.
Decision analysis allows us to select a decision
from a set of possible decision alternatives
when uncertainties regarding the future exist.
The goal is to optimize the resulting payoff in
terms of a decision criterion.
Introduction to Decision
Analysis
Gold
Junk Bond
Growth Stock
Certificate of Deposit
Stock Option Hedge
TOM BROWN
The return on each investment depends
on the (uncertain) market behavior during
the year.
Tom would build a payoff table to help
make the investment decision
S1 S2 S3 S4
D1 p11 p12 p13 p14
D2 p21 p22 p23 P24
D3 p31 p32 p33 p34
Criterion
P1
P2
P3
-100
-100
250
250
500
500
60
60
100
100
200
200
250
250
60
60
200
200
150
150
100
100
60
60
300
300
-100
-100
-200
-200
60
60
00
-150
-150
-600
-600
60
60
Minimum
Minimum
Payoff
Payoff
-100
-100
-150
-150
-600
-600
60
60
e
o
pt
i
m
al
d
ec
i
s
ion
Decisions
Decisions
Gold
Gold
Bond
Bond
Stock
Stock
C/D
C/Daccount
account
The
TheMaximin
MaximinCriterion
Criterion
Large
LargeRise
Rise Small
Smallrise
rise No
NoChange
Change Small
SmallFall
Fall Large
LargeFall
Fall
The
TheRegret
RegretTable
Table
Decision
Decision Large
Largerise
riseSmall
Smallrise
riseNo
Nochange
changeSmall
Smallfall
fall Large
Largefall
fall
600
150
00
00
60
Gold
600
150 Let us build
60
Gold
the Regret Table
250
50
50
400
210
Bond
250
50
50
400
210
Bond
00
00
100
500
660
Stock
100
500
660
Stock
440
190
140
240
00
C/D
440
190
140
240
C/D
Investing
in Stock
generates
no regret
when the
market
exhibits
a large
rise
Decision
Decision Large
Largerise
rise Small
Smallrise
rise No
Nochange
changeSmall
Smallfall
fall Large
Largefall
fall
-100
100
200
300
Gold
Investing in
gold generates
-100
100
200
300 a regret00of
Gold
250
200
150
-100
Bond
250
200 600 when
150the market
-100exhibits-150
-150
Bond
a
large
rise
500
250
100
-600
Stock
500
250
100 The -200
-200
-600
Stock
op 60
60
60
60
60
C/D
60
60
60
60
C/D
ti 60
500 (-100) = 600
al
de
ci s
io
The
Maximum
n
TheRegret
RegretTable
Table
Maximum
Decision
Decision Large
Largerise
rise Small
Smallrise
riseNo
Nochange
change Small
Smallfall
fall Large
Largefall
fall Regret
Regret
600
150
00
00
60
600
Gold
600
150
60
600
Gold
250
50
50
400
210
400
Bond
250
50
50
400
210
400
Bond
00
00
100
500
660
660
Stock
100
500
660
660
Stock
440
190
140
240
00
440
C/D
440
190
140
240
440
C/D
An aggressive decision maker looks for the decision with the highest
payoff (when payoff is profit).
op
tim
al
The Maximax Criterion
Maximum
de
c
Decision Large rise Small rise No change Small fall Largeisifall
on Payoff
-100
100
200
300
0
300
Gold
250
200
150
-100
-150
200
Bond
500
250
100
-200
-600
500
Stock
60
60
60
60
60
60
C/D
(i)
Based on the above information and under a variety of other criteria
(like maxmin, minmax etc), should the firm introduce the product?
(ii)
A consulting firm specializing in new product introduction has offered
its services to firm. Its betting average in similar situations is as follows.
(i) When advice was given (either by client firm or by others in the
market place) on products that later proved to be successful, then firm gave
GO advice eight times out of ten.
(ii) When advice was given (either by client firm or by others in the
market place) on products that later proved to be unsuccessful, then firm
gave STOP advice fifteen times out of twenty.
PAY-OFF TABLE-1
PAY-OFF TABLE-1
Successful
Unsuccessful
GO
$500
STOP
$0
.2
.8
Prior Probs.
-300
Choose the action which has the max. of these max. profits.
PAY-OFF TABLE-2
EVENTS OR STATES OF NATURE
ACTIONS
Successful
Unsuccessful
MV
GO
$500
-300
$500
STOP
$0
0
$ 0
Answer: Max-max Action = Go, Optimal Profit = $500,000
For Go Max payoff is $500
For Stop max payoff is $0
Maximum of these above
mentioned maximum values is
$500 so choice is GO decision
PAY-OFF TABLE-3
Payoff table 5
Successful
GO
$500
STOP
$0
Unsuccessful
-300
0
ACTIONS
GO
$300
STOP
$0 - -300 =
$500 - $0 = $500
$0 -$ 0 = $0
Practice Problem.
Decision
Office building
Apartment
Apartment
Apartment
Apartment
Expected
Expected Payoff
Payoff == (Probability)(Payoff)
(Probability)(Payoff)
(The summation is calculated across all the states of
nature)
Decision
Gold
Bond
Stock
C/D
Prior Prob.
eo
ptim
al d
Criterion ecisio
n
Expected
The Expected Value
Value
Large rise Small rise No change Small fall Large fall
-100
250
500
60
0.2
100
200
250
60
0.3
200
150
100
60
0.3
300
-100
-200
60
0.1
0
-150
-600
60
0.1
100
130
125
60
The-100
Expected Value of Perfect Information
Large
Decision Large rise
rise Small rise No change Small fall
Large fall
250
Gold
-100
100
200
300
0
Bond
Stock
C/D
Probab.
Stock
250
500
60
600.2
500
200
250
60
0.3
150
100
60
0.3
-100
-200
60
0.1
-150
-600
60
0.1
Decision
Gold
Bond
Stock
C/D
Probab.
The-100
Expected Value of Perfect Information
Large rise
250
-100
250
500
60
600.2
500
Small rise
100
200
250
60
0.3
No change
Small fall
200
150
100
60
0.3
Large fall
300
-100
-200
60
0.1
0
-150
-600
60
0.1
The Forecast
predicted
80%
20%
70%
30%
50%
50%
Small Fall
40%
60%
Large Fall
0%
100%
P(B Ai)P(Ai)
Posterior Probabilities
Probabilities determined
after the additional info
becomes available.
Prior probabilities
Probability estimates
determined based on
current info, before the
new info becomes available.
Nature
Large Rise
Small Rise
No Change
Small Fall
Large Fall
Prior
Prob.
0.2
0.3
0.3
0.1
0.1
Prob.
(State|Positive)
X
0.8
0.7
0.5
0.4
0
Joint
Prob.
0.16
0.21
0.15
0.04
0
Posterior
Prob.
0.286
0.375
0.268
0.071
0.000
Nature
Large Rise
Small Rise
No Change
Small Fall
Large Fall
Prior
Prob.
0.2
0.3
0.3
0.1
0.1
Prob.
(State|Positive)
X
0.8
0.7
0.5
0.4
0
Joint
Prob.
0.16
0.21
0.15
0.04
0
Posterior
Prob.
0.286
0.375
0.268
0.071
0.000
0.16
0.56
Nature
Large Rise
Small Rise
No Change
Small Fall
Large Fall
Prior
Prob.
0.2
0.3
0.3
0.1
0.1
Prob.
(State|Positive)
Joint
Prob.
0.8 = 0.16
0.7
0.21
Observe the revision in
0.5
0.15
the prior probabilities
0.4
0.04
0
0
Posterior
Prob.
0.286
0.375
0.268
0.071
0.000
Nature
Large Rise
Small Rise
No Change
Small Fall
Large Fall
Prior
Prob.
0.2
0.3
0.3
0.1
0.1
Prob.
Joint
(State|negative) Probab.
0.2
0.3
0.5
0.6
1
0.04
0.09
0.15
0.06
0.1
Posterior
Probab.
0.091
0.205
0.341
0.136
0.227
-100
250
500
60
0.286
0.091
100
200
250
60
0.375
0.205
200
150
100
60
0.268
0.341
300
-100
-200
60
0.071
0.136
EV(Invest in.
GOLD Positive forecast) =
=.286(-100 )+.375(100 )+.268( 200 )+.071(300 )+0( 0
) =
$84
EV(Invest in .
GOLD Negative forecast) =
=.091(-100 )+.205( 100 )+.341(200 )+.136(
)+.227( 0 ) =
$120
300
0
-150
-600
60
0
0.227
Decision Trees
The Payoff Table approach is useful for a
non-sequential or single stage.
Many real-world decision problems
consists of a sequence of dependent
decisions.
Decision Trees are useful in analyzing
multi-stage decision processes.
Characteristics of a decision
tree
n1
o
i
is
c
e
1
D
t
s
o
DeC
cis
Cos ion 2
t2
)
S
1
(
P
P(S2)
the consultant.
After this decision is made more decisions follow
regarding
Application for the variance.
Purchasing the option.
Purchasing the property.
su
n
co
e
0
ir
=
h
t
t
no Cos
o
Hi
re
c
Co ons
ult
st
an
=
t
-5
00
0
ing
h
t
no
Do
0
Buy land
-300,000
Pu
rch
as
-20
,00 e op
tio
0
n
Le
to t us
no c
t h on
ire sid
a er
co th
ns e d
ul ec
ta
nt isio
n
0
3
12
pro
p
A
0.4
De
nie
d
0.6
Purchase option
and
apply for variance
Build
-500,000
Buy land
-70,000
-300000 30000 + 260000 =
Sell
260,000
Build
Sell
-300,000
-500,000
950,000
100,000
-50,000
Buy land
-300,000
Build
-500,000
Buy land
-300,000
-500,000
950,000
100,000
Apply 12
for variance
-30,000
Purchase option and
apply for variance
-50,000
61
ltant
u
s
con
e
r
i
th
o
n
Do
0
co
-5 nsu
00
lta
0
nt
N
Do
ict al
d
e rov
r
P pp
A
0.
4
Hi
re
Done
ng
i
h
ot
Buy land
-300,000
Pur
cha
se
opt
-20
ion
,00
0
-5000
ict
ed
Pr nial
De
0.6
BILL GALLEN
The Decision Tree
ing
h
t
o
Do N
Buy land
-300,000
Pur
c ha
se o
ptio
-20
n
,000
-5000
Apply for variance
-30,000
ed
v
o
pr
p
A
?
De
nie
d
Sell
260,000
Co
ns
ul t
an
Sell
950,000
tp
re
di c
ts
a
ap
pr
ov
al
115,000
-75,000
Build
-500,000
Sell
950,000
Sell
260,000
115,000
-75,000
Build
-500,000
Sell
950,000
Sell
260,000
115,000
-75,000
22
23
pr
p
A
?
.7
De
nie
d
?
.3
26
Build
-500,000
24
Sell
950,000
Sell
260,000
115,000
25
-75,000
27
Hi
re
$20,000
Ap
pr
ot
n
$20,000 o
D ire
h
ove
d
$10,000
$58,000
l
rova
.3
ied
pp
a
s
ict
d
e
r
P
.4
Pre
di c
ts d
$-5,000
eni
.6
al
.7
Sell
land
Do nothing
$-75,000