UNIT – IV

INVENTORY MODELS,
SIMULATION AND
DECISION THEORY
Inventory Models – EOQ and EBQ
Models (with and without shortages)
– Quantity discount Models.
Decision making under risk – Decision
trees – Decision making under
uncertainty.
Monte – Carlo simulation

03/05/20
2
 INVENTORY CONTROL
• Inventory may be defined as the stock of
goods, commodities or other economic
resources that are stored or reserved fro
smooth and efficient running of business
affairs.

03/05/20
3
• Inventory control may be defined
as “the function of directing the
movement of goods through the
entire manufacturing cycle from the
requisitioning of raw materials to the
inventory of finished goods in an
orderly manner to meet the
objectives of maximum customer
service with minimum investment
and low cost plant operation.
03/05/20
4
 Classification
• Direct Inventories: Direct
inventories include those items which
play a direct role in the manufacture
and becomes an integral part of
finished goods, which can be further
classified into 3 groups
– Raw materials
– Work in process
– Finished goods
03/05/20
5
• In direct Inventories: Indirect
inventories include those items which
are necessary for manufacturing but
do not become an essential
component of the finished
production, such as lubricants. These
can be further classified as
– Transit or Pipeline inventories: These
are also called movement inventories.
The function of those inventories is to
cover the delays in handling and transit.
03/05/20
6
 – Buffer inventories: These inventories
are held to prevent stock out due to
uncertain demand and supply
fluctuations.
– Lot size or cycle inventories: These
inventories are hold due to the fact
that order are placed in lots rather the
purchasing the exact amount of
inventory which may be needed at the
point of time.

03/05/20
7
 – Decoupling Inventories: These
inventories serve the function of
decoupling operations in a production
system, which helps to permit the various
production activities to operate more
independently, they do not have to rely
completely on the schedule of output of
prior activities in the production process.
– Seasonal inventories: The function of
these inventories is to meet the seasonal
fluctuations in demand economically

03/05/20
8
 – Fluctuation Inventories: In real-life
problems, there are fluctuations in the
demand and lead times that affect the
production of the items.
– Anticipation inventories: These are
built up in advance for the season of
large scales, a promotion programme
or a plant shut down period.
Anticipated inventories keep men and
machine for future participation

03/05/20
9
 Phases of Inventory
management:
• An inventory management system can be
described as a coordinated set of rules and
procedure which are helpful in taking
decisions such as – when to order and how
much to order each item of inventory
required in the manufacturing or
procurement process, to meet customers
demand.
– Assessment of inventory requirement:
– Ensuring material quality
– Timely inventory flow and inventory control
– Investment in inventories
– Transactions maintenance.
03/05/20
10
 Objectives of Inventory
Management
• to minimize the financial investment in inventories
• to ensure that the value of materials consumed is
minimum
• to maintain timely records of inventories
• to provide scientific base for short term and long
range planning of inventory requirements
• To ensure timely action for replenishment.
• to protect the bank of inventories
• to standardize and centralize information on stock
levels and progress of stock issues
• to provide a safeguard for variations in raw
materials delivery time
• to allow flexibility in production scheduling
• to reduce surplus stock
03/05/20
11
 Costs involved in Inventory
Problems
• Carrying Cost (OR) Holding cost(:C1
The cost associated with carrying or
holding the goods in stock is known
as holding cost or carrying cost per
unit per unit of time.
Holding cost is assumed to directly
vary with the size of inventory as
well as the time the item is held in
stock.
03/05/20
12
• If P is the purchase price at an item
and I is the stock holding cost per
unit time, then the holding
C1 cost
=I*P

03/05/20
13
• Shortage Cost (OR) Stock out C2
Cost
The penalty costs that are incurred
as a result of running out of stock
(ie., shortage) are known as shortage
or stock out costs.
• Purchase or Production Cost: The
cost of purchasing a unit of item is
known as purchase cost
03/05/20
14
• Set- up CostC3
These costs are associated with
obtaining goods through placing an
order or purchasing or manufacturing
or setting up machinery before starting
production. So they include costs of
purchase, requisition, follow up
receiving the goods, quality control,
etc. These are called ordering costs or
replenishment costs, or set up cost.

03/05/20
15
 Definitions
• Lead-time: The elapsed time between the
placement of the order and its receipts in inventory
is known as lead-time Or The time gap between
placing of an order and its actual arrival in the
inventory.
• Reorder level (R.O.L): This is the time when we
should place order by taking into consideration the
interval between placing the order and receiving
the supply.
• Economic Order Quantity (E.O.Q): Economic
order quantity is that size of order, which minimizes
total annual cost of carrying inventory and the cost
of ordering under the assumed conditions of
certainty and that annual demands are known.
03/05/20
16
 Deterministic inventory
models:
• EOQ with no shortage
– Model I: Purchasing model with no
shortages
– Model II: Manufacturing model with no
shortages

03/05/20
17
• EOQ with shortage
– Model III: Purchasing model with
shortages
– Model IV: Manufacturing model with
shortages.

03/05/20
18
 Model I: Purchasing model
with no shortages
• q = Quantity of material ordered at
each order point (units per order)
• C1 = Cost of carrying (holding cost)
one unit in inventory for one year.
• C0 = Average sot of completing on
order for material
• TIC = total annual stock costs for a
material

03/05/20
19
• Demand rate is uniform and
production rate is infinite
– A manufacturer has to order supply
goods at a uniform rate of R per unit
time. He starts a production run for
every time ‘t’ units where t is fixed and
set up cost is C3. If the productions are
made at intervals t, then the optimum
quantity to be produced during each
production run is
03/05/20
20
 2C3 R
Economic lot size q
C1
2C3
Time interval t
C1 R

Mean Average Cost C  2C1C3 R

03/05/20
21
 Model II: Manufacturing
Model with no shortages
• Demand rate is uniform and
production rate is infinite
K 2C3 R
q
K R C1
K 2C3
t 
K R C1 R

03/05/20
 K R
C 2C1C3 R
K

03/05/20
23
 Problems
1. The annual demand for an item is 3200 units. The
unit cost is Rs 6 and inventory carrying charges is
25% per annum. If the cost of 1 procurement is Rs
150/-. Determine
1. Optimal lot size
2. Time between consecutive orders
3. No of orders per year
4. Optimal total cost

03/05/20
• Given R = 3200
• C1= 25 % (6) =1.5
• C3 = Rs 150

2C3 R
• EOQ q 
C1
=800

03/05/20
25
• T=q/R = 800/3200 = 1/4th of a year
= 3 months

• Number of orders = 1/T = 4

• Total Optimal Cost = 20,400

C  2C1C3 R  R * Pr ice

03/05/20
26
2. A shopkeeper has uniform demand of an item at
the rate of 600 per year. He buys from a supplier at
the cost of Rs 8 per item and the cost of ordering is
12 each item. If the stock holding cost are 20% per
year of the stock value. How frequently should he
replenish his stock and what is the optimal lot size.

03/05/20
27
3. A company rivets at the rate of 5000 Kg per year.
Rivets costing Rs 2 per Kg. If it costs Rs 20/- to
place an order and carrying cost of inventory is 10
% per year. How frequently the order is placed and
how much.
4. A certain item costs Rs 235 per tonne. The monthly
requirements are 5 tonne and each time the stock is
replenished. There is a set up cost of Rs 1000. The
cost of
03/05/20
28
Carrying inventory has been estimated at 10% of the
average inventory per year. What is the optimum
order quantity.
5. A contractor has to supply bearings 10,000 to an
automobile manufacture. He find that when he
starts a production, he can produce 25,000
bearings a month. The cost of holding bearing in
stock for one year is Rs 2 and set up cost of

03/05/20
29
Production run is Rs 180. How frequently should the
production be made.
6. An item is produced at the rate of 50 per day. The
demand occurs at the rate of 25 per day. If the set
up cost is Rs 100 and holding cost is 1 paise per
unit. Find the economic lot size, time of the cycle
and minimum total cost per run.

03/05/20
30
7. A manufacturing company purchases 9000
parts of a machine for its annual
requirements , ordering one month usage
of time. Each part costs Rs 20/-. The
ordering cost per order is Rs.15/- and the
carrying charges are 15% of the average
inventory per year. You have been asked to
suggest a more economical purchasing
policy for the company. What advice would
you offer, and how much would it save the
company per year?
03/05/20
31
8. An aircraft company uses rivets at an
approx customer rate of 2500 kg per
year. Each unit costs Rs 30 per kg and
the company personnel estimate that it
costs Rs 130 to place an order, and
that the carrying cost of inventory is
10% per year. How frequently should
orders for rivets be place? Also
determine the optimum size of each
order.
03/05/20
32
 Model III: Purchasing model
with shortages
(Demand Rate uniform production rate
is infinite, shortages allowed)
C1  C2 2C3 R
q
C2 C1
q
C0 
C1
2C1C3 R t
C1  C2
R
C1 2C3 R
Im 
C1  C2 C1
03/05/20
33
 Model IV: Manufacturing
Model with shortages
• (Demand rate uniform, production
rate finite, shortages allowed)
K C1  C2 2C3 R
q
K R C2 C1
K R C2
C0  2C1C3 R
K C1  C2
K R C2 2C3 R
Im 
K C1  C2 C1
03/05/20
34
 Problems
1. The demand for an item is 18,000
units per year. The holding cost per
unit time is Rs 1.20 and the cost of
shortage is Rs 5, the production cost
is Rs 400. Assuming that
replenishment rate is instantaneous,
determine the optimal order
quantity.

03/05/20
35
2. The demand for a certain item is 16
units per period. Unsatisfied demand
causes shortage cost Rs 0.75 per unit
per short period. The cost of initiating
purchasing action is Rs 15 per
purchase and the holding cost is 15%
of average inventory valuation per
period. Item costs Rs 8 per unit. Find
the minimum cost purchase quantity.

03/05/20
36
3. A commodity is to be supplied at a constant rate
of 200 units per day. Supplies for any amounts
can be had at any required time, but each
ordering costs Rs 50. Cost of holding the
commodity in inventory is Rs 2 per unit per day
while the delay in the supply of the items induces
a penalty of Rs 10 per unit per delay of one day.
Formulate the average cost function of this
situation and find the optimal policy (q, t) where t
is the reorder cycle period and q is the inventory
level after re-order. What should be the best
policy if the penalty cost becomes infinite.

03/05/20
37
4. The demand for an item in a company is
18,000 units per year, and the company
can produce the item at a rate of 3000 per
month. The cost of one set up is Rs 500
and the holding cost of one unit per month
is 15 paise. The shortage cost of one unit is
Rs 20 per month. Determine the optimum
manufacturing quantity and the number of
shortages. Also redefine the manufacturing
time and time between set-ups.

03/05/20
38
5. A contractor undertakes to supply diesel engines
to a truck manufacturer at the rate of Rs 25 per
day. There is a clause in the contract penalising
him Rs 10 per engine per day late for missing the
schedule delivery rate. He finds that the cost of
holding a completed engine in stock Rs 16 per
month. His production process is such that each
month he starts a batch of engines through the
shops, and all these engines are available for
delivery any time after the end of the month.
What should his inventory level be at the
beginning of each month?

03/05/20
39
 Quantity Discount Model
• Inventory Model with Price Breaks
– In this model the production or purchase
cost per unit is a variable. This depends
on the quantity manufactured or
purchased. This usually happens when
discounts are offered for the purchase of
large quantities. These discounts take
the form of Price-Breaks

03/05/20
40
• Consider the following three cases,
whereC (q )  2C KPR  KR  1 C p
0 3 3
2
2C3 R
q
KP
C3 R 1
C (q)   qPI  PR
q 2
K- purchasing cost of care unit
p – holding cost /month expressed as fraction
of the value of the unit
03/05/20
41
 Case (i): Single break up
• Given: Unit purchasing
Range
cost of quantity
k11 0  q1  b1
k12 q2  b1
b  q2 c2 (b)  c0 (q1 ) q1
(i)If and c(q )  c0,the
(q1 ) optimal lot size is
and bminimum
 q2 (b)  c0of
c2value (q1 ) b
(ii)If and c(q )  c2 (b) , the optimal lot size is
and min b  qvalues
2
q2 c ( q )  c 0 ( q2 )
(iii) If , the optimal lot size is and min
03/05/20
42
 Case (ii): 2 break-ups
• Unite Purchasing Cost Range of quantity
k11 0  q1  b1
k12 b1  q2  b2
k13 b2  q3

( i ) Calculate q3. If q3 > b2, optimal purchase quantity is q3

(ii) If q3 ≤ b2, calculate q2 since q3<b2, then necessarily q2<b2.
As a consequence we have q2<b1 or q2> b1.
(iii) If q3<b2 and b1< q2 < b2, compare c0 (q2 ) with .
c3 (b2 )
The smaller of these quantities will be optimal purchase
quantity.
(iv) If q3<b2 and q2<b1, calculate c3(q1) which will necessarily
satisfy the inequality q1<b1.

03/05/20
43
 Case (iii): ‘n’ break-ups
• Unite Purchasing Cost Range of
quantity
k11 0  q1  b1
k12 b1  q2  b2
... ...
k13 b2  q3
03/05/20
44
(i) Calculate qn. Ifqn>bn-1, optimal
purchase quantity is qn
(ii) If qn<bn-1, calculate qn-1. If qn-1bn-2
proceed as in the case of one price break
(iii) If qn-1<bn-2, computeqn-2. If qn-2 bn-
3, proceed as in the case of 2 price
breaks
(iv)If qn-2<bn-3 compute qn-3. If qn-3bn-4
compare c0(qn-3) with c(bn-3) c(bn-2)
and c(bn-1)
03/05/20
45
 Problems
1. Find the optimal order quantity for a product
for which the price-break is as follows:
Quantity Unit. Cost
0≤Q1<50 Rs.10
50≤Q2<100Rs. 9
100 ≤Q3 Rs. 8
The monthly demand for the product is 200
units, the cost of the storage is 25% of the
unit cost and ordering cost is Rs 20.00 per
order

03/05/20
46
• Given R = 200, P= 0.25, C3= Rs 20.00

2C3 R
• Q3 = = 63 Units
KP
• 63 units < 100 (i.e) Q3 < b2, we
compute Q2
2C3 R
• Q2 = = 60 Units
KP
03/05/20
47
• Now Q2 > b1 (50 Units), the optimum
purchase quantity is determined

C3 R 1
• C(Q2) =
 qPk  PR
q 2
= 20 * 200 1
 * 60 * 9 * 0.25  9 * 200
60 2
= Rs 1934.16
03/05/20
48
 C3 R 1
• C(b2) =  qPI  PR
q 2
20
= * 200 1
 *100 *8* 0.25  8* 200
100 2
= Rs 1740.00

Therefore, the optimum purchase

quantity is 100 Units
03/05/20
49
2. Find the optimal order quantity for which
the price break-up is as follows:
Quantity Unit cost
0≤q1<500 Rs 10
500≤q2<750 Rs 9.25
750 ≤q3 Rs 8.75
The monthly demand for the product is 200
units, shortage cost is 2% of the unit cost
and the cost of ordering is Rs 100.

03/05/20
50
3. Find the optimum order quantity for a product
for which the price breaks are as follows:
Quantity Purchasing Cost
0≤q1<100 Rs 20
100≤q2<200Rs 18
200<q3 Rs 16
The monthly demand for the product is 400. The
storage cost is 20% of the unit cost of the
product and the cost of ordering is Rs 25 per
month

03/05/20
51
4. Find the optimal quantity for a product where
the annual demand for the product is 500 units.
The cost of storage per unit per year is 10% of
the unit cost and the ordering cost is per order is
Rs180. The unit costs are given below
Quantity Unit Cost
0≤q1<500 Rs 25
500≤q2<1500 Rs 24.80
1500<q3<3000 Rs 24.60
3000<q4 Rs 24.40

03/05/20
52
5. Annual demand for an item is 500 units, ordering
cost is Rs 18. Inventory carrying cost is 15 per
year per unit.
Quantity Unit Price
1-15 10
16-149 9
150-549 8.75
550 and over 8.50
Specify optimal order quantity and the
corresponding price of this item

03/05/20
53
 Simulation
Simulation is the process of designing a
model of a real system and conducting
experiments with this model for the
purpose of understanding the behavior for
the operation of system.
OR
Simulation is a representation of reality
through the use of a model or other device
which will reach in the same manner as
reality under a given set of conditions.
03/05/20
54
 Examples
• Testing an aircraft model in a wind
tunnel
• Children cycling park with various
signals and crossing – to model a
traffic system
• Planetarium – to determine the
behavior of a real system in true
environment.

03/05/20
55
 Types
• Analogue Simulation
• Computer Simulation.

03/05/20
56
 Steps of simulation process
• Identify the problem
• Identify the decision variables and
decide the performance criteria and
decision rules
• Construct a numerical model
• Validate the model
• Design the experiment
• Run the simulation model
• Examine the results
03/05/20
57
 Advantages and
disadvantages of simulation
• Advantages
– This approach is suitable to analyze large
and complex real-life problems which cannot
be solved by usual quantitative methods
– It is useful for sensitivity analysis of complex
systems.
– Simulation cannot be used to ‘experiment’
on a model of a real situation without
incurring the cost of operating on the
systems

03/05/20
58
 – Simulation can be used as a pre-service test to try out
new policies and decision rules for operating a system.
Disadvantages.
– In some case simulation models are expensive
– It is the trial and error approach
– Each application of simulation is adhoc to a
great extent.
– Simulation model does not produce answers by
itself. The user has to provide all the
constraints for the solutions which he wants to
examine.

03/05/20
59
 Limitations
• Quantifications of the variable may be
difficult.
• Large number of variable makes
simulation unwieldy and more difficult.
• Simulation may not yield optimum
result.
• Simulation may not be cheap.
• Simulation may not be less time
consuming.
03/05/20
60
 Monte-Carlo simulation
• The Monte – Carlo method is a
simulation technique in which the
statistical distributions functions are
created using a series of random
number.
• This method is generally used to solve
problems, which cannot be adequately
represented by mathematical model or
where the solution cannot be arrived
by analytical method.

03/05/20
61
 Procedure
• Setting up a probability distribution
for variables to be analyzed
• Building a cumulative probability
distribution for each random variable
• Generate random numbers. Assign
an appropriate set of random
numbers to represent value for each
random variable.

03/05/20
62
• Conduct the simulation experiment
by means of random sampling
• Repeat the previous step until the
required number of simulation runs
has been generated
• Design and implement a course of
action and maintain the control.

03/05/20
63
 Problems
1. Customers arrive at a milk booth for
the required service. Assume that
the inter-arrival time and service
time are constants and given by 1.5
and 4 minutes. Simulate the system
by hand computation for 14 minutes.
1. What is the waiting time per customer?
2. What is the percentage idle time for the
facility?

03/05/20
64
2. A bakery keeps stocks of a popular
brand of cake. Pervious experience
shows the following demand pattern for
the item with associated probabilities.

Daily Demand
(number) 0 10 20 30 40 50

Probability 0.01 0.2 0.15 0.5 0.12 0.02

03/05/20
65
Use the following sequence of random
numbers to simulate the demand for
next 10 days.
Random number 25 39 65
76 12 05 73
89 19 49.
Also estimate the daily average
demand for the cakes on the basis of
simulated data
03/05/20
66
3. A tourist car operator finds that during the past few months the car’s use
has varied so much that the cost of maintaining the car has varied
considerably. During the past 200 days the demand for the car fluctuation is
given below

Trips / Week 0 1 2 3 4 5
• Use the random numbers 82 95 18 96 20 84 56
11 Frequency
52 03 16 the24
to simulate demand30
for a 1060 40 30
week period.

03/05/20
67
4. A company manufacturers 30 items per day. The
sale of these items depends upon demand which
has the following distribution.
Sales 27 28 29 30 31 32
Prob 0.1 0.15 0.2 0.35 0.15 0.05
The production cost and selling price of each units are
Rs 40/- and Rs 50/-. Any unsold product is to be
disposed off at a loss of Rs 15/- per unit. There is a
penalty of Rs 5 per unit if the demand is not met.
Use the random numbers 10 99 65 99
95 01 79 11 16 20 to estimate the
profit/loss for the company for the next 10 days
03/05/20
68
5. The director of finance, for a farm cooperative is
concerned about the yields per acre, she can
expect from this year’s corn crop. The probability
distribution of the yields for the current weather
conditions is given below:

Yield in Kg 120 140 160 180

Prob 0.18 0.26 0.44 0.12
She would like to see the simulation of the yields She
might expect over the next 10 years for weather
conditions
03/05/20
69
Simulate the average yield she might expect per
acre using the random numbers 20 72 34
54 30 22 48 74 76 02.
She is also interested in the effect of market price
fluctuations on the cooperative’s farm revenue.
She makes this estimate of per kg prices for corn
Price per
kg 2 2.1 2.2 2.3 2.4 2.5
Prob 0.05 0.15 0.30 0.25 0.15 0.10

03/05/20
70
• Simulate the price she might expect
to observe over the next 10 yeas
using the following random numbers
82 95 18 96 20 84 56 11 52
03

03/05/20
71
6. A company manufactures around
200 mopeds. Depending upon the
availability of raw materials and
other conditions, the daily production
has been varying from 196 mopeds
to 204 mopeds, whose probability
distribution is given below:

03/05/20
72
Prod/
Day 196 197 198 199 200 201 202 203 204
 Prob  0.05 0.09 0.12 0.14 0.20 0.15 0.11 0.08 0.06

• The finished mopeds are transported in a specially

designed three storeyed lorry than can accommodate
only 200 mopeds. Using the following random numbers
82 89 78 24 53 61 18 45 04
23 50 77 27 54 10 simulate the
process to find out
• What will be the average number of mopeds waiting in
the factory?
• What will be the average number of empty space on the
lorry?

03/05/20
73
7. An airline has 15 flights leaving a base per day,
each with hostess. The airline keeps three
hostesses in reserve so that they may be called in
case the scheduled hostess for a flight reports sick.
The probability distribution for daily number of sick
hostess is as follows
No. of sick 0 1 2 3 4 5
Prob 0.2 0.25 0.2 0.15 0.1 0.1
Use Monte – Carlo method to estimate the utilization of
reserve hostesses and also the probability that at least
one flight will be cancelled in a day because of non-
availability of hostess.
03/05/20
74
 DECISION THEORY
• Decision theory or decision analysis
is an analytical and systematic
approach to decision making where
the decision maker has several
feasible and viable decision
alternatives from which he or she has
to select the best alternative on the
basis of some standards decided in
advance
03/05/20
75
 Components of a Problem

– Decision Maker
– Objectives
– System Environment
– Alternate course of Action
– Unequal efficiencies of chances

03/05/20
76
 Types of Problems
• Deterministic Problems
• Stochastic Problems
• Decision making under uncertainty
• Decision making under risk

03/05/20
77
 Deterministic Problems
• These are the problems in which
each course of action results is only
one outcome. For such problems the
probability of each outcome is
known. The measure of performance
used in deterministic problems is
that of the maximization of utility.

03/05/20
78
 Stochastic Problems
• These problems are those in which
each course of action can result in
alternative outcomes, the
probabilities of which are known or
can be estimated. This criterion that
is generally accepted as the most
appropriate for such problems is
that of maximizing the expected
utility.
03/05/20
79
 Decision making under
uncertainty
• A situation where the decision maker is
unable to assess the probability of any state
of nature is referred to as decision making
under uncertainty.
• The decision making is based on several
criteria
– Laplace criterion
– Maximin or minimax criterion
– Maximax or minimin criteriion
– Hurwicz criterion
– Regret criterion
03/05/20
80
 Laplace criterion
• This criterion is based on the simple
principle that various events can be
treated equally likely. Under this
assumption the expected payoff for
each act is computed, followed be
mean of these expected payoff.

03/05/20
81
 Maximin or minimax
criterion
Maximin criterion is a conservative
approach to decision making. The
decision maker tries to avoid the
worst choice. Here the minimum
payoff over the various events or
states of nature is determined by the
decision maker and an act is selected
for which the minimum payoff is the
highest.
03/05/20
82
• Minimax criterion is used by a
decision maker when consequences
are given in the form of cost or
opportunity loss. In this approach the
decision maker determines the
maximum cost over all the states of
nature and then selects the act for
which the cost is minimum.

03/05/20
83
 Maximax or minimin
criterion
• Maximax criterion is an optimistic
approach where a decision maker
determines the maximum payoff for
each act and then an act is selected
which provides the highest returns.
In other words a decision maker
selects the act which gives the
overall maximum among the
maximum payoffs
03/05/20
84
• Minimim criterion is used when the
consequences are given in the form
of cost or opportunity loss. According
to his approach, the decision maker
determines the minimum opportunity
loss or cost for each act and selects
one which provides overall minimum
cost or opportunity loss.

03/05/20
85
 Hurwicz Criterion
• This approach focuses on a more
poised selection of alternative by
opting for neither a completely
pessimistic approach nor a
completely optimistic approach.
Hurwics introduced a coefficient of
optimism, denoted by , which varies
on a scale ranging form 0 to 1.

03/05/20
86
In this scale 0 indicates an extremely
pessimistic approach to the future
and 1 indicates an extremely
optimistic approach to the future.
Hurwicz criterion value =  (Max
value) + (1-)(Min Value)

03/05/20
87
 Regret criterion
• In this criterion, a decision maker
selects the course of action that
minimizes the maximum regret. First
the given payoff is converted into an
opportunity loss or regret matrix, then
the decision maker determines the
maximum regret for each act and then
select the overall minimum regret
value from the list of maximum regret
values.
03/05/20
88
 Decision making under risk
• Decision making under risk is a situation
where more than one state of nature
exists and the decision maker has
sufficient information to assign
probability values to the likelihood of
occurrence of each of these states.
• The three approaches a decision maker
uses to evaluate various courses of
action and select the best course of
action are
03/05/20
89
• Expected monetary value (EMV)
• Expected opportunity loss (EOL)
• Expected value of perfect
information (EVPI)

03/05/20
90
 Expected Monetary Value
(EMV)
• EMV is the sum of payoffs for each
course of action multiplied by the
probabilities associated with each
state of nature.

k
EMV   xij p j
j 1

03/05/20
91
 Expected Opportunity Loss
(EOL)
• EOL criterion is another approach
based on which a decision can be
taken. The EOL can be computed as
k
EOL   lij p j
j 1

• Where l is the opportunity loss

associated with each act

03/05/20
92
 Expected Value of Perfect
Information (EVPI)
• EVPI is referred to as the difference
between the expected payoff with
perfect infromation and the
maximum expected payoff computer
under uncertainty

03/05/20
93
 Problems
1. A company is faced with the
problem of a decline in its sales
turnover. To overcome this problem,
it has decided to opt for any of the
four strategies: heavy advertisement
(S1); increase in number of sales
executives (S2); adding new features
to the products (S3) and increasing
the price of the product (S4).
03/05/20
94
Of these 4 acts, there may be 4
possible states of nature which are
40% increase in sales (E1); a 30%
increase in sales(E2); a 25% increase
in sales(E3) and a 22% increase in
sales (E4). The company executives
have worked out the yearly net profit
that would result if any of the 4
strategies are selected.
03/05/20
95
 State Acts
of S1 S2 S3 S4
Natur
e

E1 100 250 850 500

E2 200 500 300 700
E3 400 600 600 200
On the basis
E4 of the five criteria for
600 800 350 500
decision making under uncertainty,
suggest which act should be adopted
by the decision maker
03/05/20
96
 Solution
(i) Laplace Criterion
Act Mean Expected Payoff
S1 (100+200+400+600)/4=325
S2 (250+500+600+800)/4=537.5
S3 (850+300+600+350)/4=525
S4 (500+700+200+500)/4=475

03/05/20
97
(ii) Maximin criteria
Act S1 S2 S3 S4
Minimum 100 250 300 200

Here, the maximum is under S3.

Therefore, the decision maker will
select act S3

03/05/20
98
(iv) Hurwicz Criterion
Act Min Max Hurwicz Value

S1 100 600 600*.6+100*.4=400

S2 250 800 800*.6+250*.4=580
S3 300 850 850*.6+300*.4=630
S4 200 700 700*.6+200*.4=500

According to Hurwicz, a decision

maker will select act S3

03/05/20
99
 (v) Regret criterion
State Act
of S1 S2 S3 S4
Natur
e

E1 850-100=750 850- 850-850=0 850-

250=600 500=350
E2 700-200=500 700- 700-300=400 700-700=0
500=200
E3The 600-400=200
maximum regret
600-600=0values is under
600-600=0 600-act S2
200=400
E4
(minimum)=800.
800-600=200
Therefore,
800-800=0
S2 is selected
800-350=450 800-
500=300
03/05/20
100
2. On the basis of expected monetary
value criterion, what decision should
be taken by the decision maker.
State Prob Acts
of
Natur S1 S2 S3 S4
e

E1 0.15 100 250 850 500

E2 0.30 200 500 300 700
E3 0.35 400 600 600 200
03/05/20 E4 0.20 600 800 350 500
101
• EMV for S1 = 0.15*100 + 0.3 * 200 + 0.35* 400 + 0.2 * 600
• = 470
• EMV for S2 = 0.15*250 + 0.3 * 500 + 0.35* 600 + 0.2 * 800
• = 555
• EMV for S3 = 0.15*850 + 0.3 * 300 + 0.35* 600 + 0.2 * 350
• = 497.5
• EMV for S4 = 0.15*500 + 0.3 * 700 + 0.35* 200 + 0.2 * 500
• = 455

• Maximum is at S2

03/05/20
102
3. On the basis of EOL criterion, what
decision should be taken by the
decisionState
maker.
Proba Acts
of bility
Natur S1 S2 S3 S4
e

E1 0.15 100 250 850 500

E2 0.30 200 500 300 700
E3 0.35 400 600 600 200
E4 0.20 600 800 350 500
03/05/20
103
• EMV for S1 = 0.15*100 + 0.3 * 200 + 0.35* 400 + 0.2 * 600
• = 470
• EMV for S2 = 0.15*250 + 0.3 * 500 + 0.35* 600 + 0.2 * 800
• = 555
• EMV for S3 = 0.15*850 + 0.3 * 300 + 0.35* 600 + 0.2 * 350
• = 497.5
• EMV for S4 = 0.15*500 + 0.3 * 700 + 0.35* 200 + 0.2 * 500
• = 455

• Minimum is for S4
03/05/20
104
4. Calculate the expected EVPI
State Prob Acts
of
Natur S1 S2 S3 S4
e

E1 0.15 100 250 850 500

E2 0.30 200 500 300 700
E3 0.35 400 600 600 200
E4 0.20 600 800 350 500

03/05/20
105
03/05/20
106
 Decision Trees
• A decision tree is a graphical model
of a decision process. Decision tree
allow visualizing the decision
problem and resemble drawing of
trees.
• A decision trees consists of nodes,
branches, probability estimates and
payoffs.

03/05/20
107
• Nodes can be of two types
– A decision node is generally represented
by a square and a decision point where
a decision maker must select one action
among several possible actions
– A chance node is represented by a circle
and indicates a point where the decision
maker will discover the response to his
decision.

03/05/20
108
• Branches are also two types
– A branch leading from a decision node
represents a course of action or
strategy, and a branch leading away
from a chance node represents the
nature of state.

03/05/20
109
 Problems
1. Company ABC has developed a new line
of products. Top management is
attempting to decided on the appropriate
marketing and production strategy. Three
strategies A, B and C are being
considered. Managements bets estimate
of net profit is each case is given in the
following pay-off table

03/05/20
110
Managements best estimate of the
probabilities of a strong or a weak
market are 0.45 and 0.55
respectively. Draw a decision tree to
represent the strategies of different
products?

03/05/20
111
 S 0.45
30

W
-8
0.55

A S
20
B
W
C
7
S 5

W
15

03/05/20
112
2. Mr. X had to decided whether or not to drill a well
in his farm. In his village, only 40% of wells
drilled were successful at 250 feet of depth. Some
of the farmers who did not get water at 250 feet
drilled further up to 300 feet, but only 10% struck
water at 300 feet. Cost of drilling is Rs 60/- per
feet. Mr. X has estimated that he would pay Rs
20,000 during a 5 year period in his present value
terms. If he continues to bay water from
neighbour, rather than go for a well which you
would have life for 5 years. Frame a decision tree
to the above problem.

03/05/20
113
 Do not drill
Rs 20,000

Water struck(0.4) Rs 15,000

Water struck(0.1)
Rs 18,000
Drills up to 300 f
Drills up to
250f
Rs 38,000
No water(0.9)

No water(0.6)

Rs 35,000
Do no drill

03/05/20
114
 Problems
3. A company engaged in the process of
launching a new product. The top
management of the company has three
options in terms of launching the
product in three sales zones: North (N),
South (S) and East Region (E). The
management has decided to take the
final decision on the basis of the
demand for the product, which is
divided into three categories:
03/05/20
115
• High, medium and low demand. On the
basis of past data and management’s view,
the respective probabilities fro three
categories of demand are estimated to be
0.45, 0.35 and 0.20. The table indicates
the estimated profit for various
combinations of events and acts. Construct
a decision tree and using expected value
criterion, select the alternative with the
highest expected payoff.

03/05/20
116
Demand Probabilit Course of action
y
N S E
HIGH
0.45 100 -60 -
120
MEDIUM
0.35 80 -20 40
LOW
0.20 60 150 135

03/05/20
117