Professional Documents
Culture Documents
QUANTITATIVE TECHNIQUES
Department of Distance and Continuing Education DEPARTMENT OF DISTANCE AND CONTINUING EDUCATION
University of Delhi UNIVERSITY OF DELHI
AS PER THE UGCF-2022 AND NATIONAL EDUCATION POLICY 2020
QUANTITATIVE TECHNIQUES FOR MANAGEMENT
Editorial Board
Dr. Rishi Rajan Sahay
Dr. Gurjeet Kaur
Content Writers
Dr. Mona Verma, Dr. Abhishek Kumar Singh, Dr. Sandeep Mishra,
Dr. Shubham Agarwal, Mr. Jigmet Wangdus
Academic Coordinator
Mr. Deekshant Awasthi
Published by:
Department of Distance and Continuing Education
Campus of Open Learning/School of Open Learning,
University of Delhi, Delhi-110007
Printed by:
School of Open Learning, University of Delhi
QUANTITATIVE TECHNIQUES FOR MANAGEMENT
Printed at: Taxmann Publications Pvt. Ltd., 21/35, West Punjabi Bagh
New Delhi - 110026 (500 Copies, 2024)
PAGE
Lesson 1: Linear Programming
1.1 Learning Objectives 1
1.2 Introduction 2
1.3 Linear Programming Problem 2
1.4 Formulation of Linear Programming Problem 3
1.5 Graphical Method 7
1.6 Special Cases in Graphical Method 10
1.7 Simplex Method 12
1.8 Duality 19
1.9 Sensitivity Analysis 21
1.10 Summary 26
1.11 Answers to In-Text Questions 26
1.12 Self-Assessment Questions 27
1.13 References 30
1.14 Suggested Reading 30
PAGE i
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE
Lesson 3: Transportation Problem
3.1 Learning Objectives 57
3.2 Introduction 58
3.3 Transportation Problem 58
3.4 Some Special Codes of TP 75
3.5 Summary 83
3.6 Answers to In-Text Questions 84
3.7 Self-Assessment Questions 85
3.8 References 87
3.9 Suggested Readings 87
PAGE
5.7 Summary 141
5.8 Answers to In-Text Questions 142
5.9 Self-Assessment Questions 142
5.10 References 142
5.11 Suggested Reading 142
PAGE iii
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE
7.8 Game without Saddle Point 171
7.9 Solution of 2 × 2 Games without Saddle Point 174
7.10 Dominance Rule 177
7.11 Graphical Method to Solve 2 × n, m × 2 and 2 × 2 Games 180
7.12 Applications of Game Theory 184
7.13 Limitations of Game Theory 184
7.14 Summary 185
7.15 Answers to In-Text Questions 185
7.16 Self-Assessment Questions 185
7.17 References 186
7.18 Suggested Readings 186
Glossary 187
iv PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
1
Linear Programming
Dr. Mona Verma
Assistant Professor
Shaheed Sukhdev College of Business Studies
University of Delhi
Email-Id: monavermag@sscbsdu.ac.in
STRUCTURE
1.1 Learning Objectives
1.2 Introduction
1.3 Linear Programming Problem
1.4 Formulation of Linear Programming Problem
1.5 Graphical Method
1.6 Special Cases in Graphical Method
1.7 Simplex Method
1.8 Duality
1.9 Sensitivity Analysis
1.10 Summary
1.11 Answers to In-Text Questions
1.12 Self-Assessment Questions
1.13 References
1.14 Suggested Reading
PAGE 1
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
1.2 Introduction
Linear Programming problem is a mathematical technique to optimise
the performance in terms of profit or cost under a set of constraints like
man power, machine hours, raw materials etc. specified by the company.
The technique is very powerful and found especially useful because
of its application to many different types of real business problems in
areas like finance, production, sales and distribution, personnel, mar-
keting and many more areas of management. These will be explained
in the next chapter.
As the name suggests, linear programming model consists of linear
objectives and linear constraints, which means that the variables in a
model have a proportionate relationship. For example, an increase in
manpower resource will result in an increase in work output.
2 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
Notes
1.4.1 Defining the Decision Variable and Constraints
In a given problem situation, defining the key decision variables is
important. Identifying these variables help us to develop the model.
For example, consider a manufacturer who is manufacturing three prod-
ucts A, B and C using two machines, I and II. Each unit of product A
takes 2 minutes on machine I and 5 minutes on machine II. Product B
takes 1 minute on machine I and 3 minutes on machine II. Similarly,
product C takes 4 minutes and 6 minutes on machine I and machine II,
respectively. The total available time on machine I and machine II are
200 hours and 420 hours, respectively. Each unit of A yields a profit of
Rs. 3.00, B yields Rs. 4.00 and C yields Rs. 5.00. What should be level
of production of products A, B and C that should be manufactured by
the company so as to maximize the profit?
The decision variables, objectives and constraints are identified from the
problem.
The company is manufacturing three products A, B and C. Let A be x1,
B be x2 and C be x3. x1, x2 and x3 are the three decision variables in the
problem. The objective is to maximize the profits. Therefore, the problem
is to maximize the profit, i.e., to know how many units of x1, x2 and x3
are to be manufactured.
There are two machines available, machine I and machine II with total
machine hours available as 200 hours and 420 hours. The machine hours
are the resource constraints, i.e., the machines cannot be used more than
the given number of hours.
To summarize,
Key decision: How many units of x1, x2 and x3 are to be manufactured
Decision variables: x1, x2 and x3
Objective: To maximize profit
Constraint: Machine hours
Feasibility Condition: All decision variables are non-negative.
Formulation of linear programming is the representation of problem sit-
uation in mathematical form. It involves well defined decision variables,
with an objective function and set of constraints.
4 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
To formulate the above problem, we convert the word problem into the Notes
tabular form as below:
ª ª a11 a1n º º
«« »»
«« »»
«¬ «¬ am1 amn »¼ »¼
For the above example the technological coefficient matrix can be written as
ª2 1 4º
«5 3 4»
¬ ¼
Resource Availability (bi ): The constant bi is the amount of resource i
available during the given period. A generalised format of the resource
availability matrix is:
PAGE 5
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes ª b1 º
«b »
« 2»
«#»
« »
¬bm ¼
For the above question we have b1 = 200 hours and b2 = 420 hours
Set of Constraints: A constraint is a restriction on the total amount of
particular resource required to carry out different activities
A generalised form of constraints can be written as:
a x + a x + ... + a x d (t, =) b
11 1 12 2 1n n 1
Any linear programming problem having multiple optimal solutions will Notes
have an infinite number of solutions, each with the same optimal value
of the objective function.
For any LPP, the objective function value can be increased or decreased
infinitely without any limitations. Such solution is known as unbounded
solution.
PAGE 7
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
8 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
7KH LQHTXDOLW\ FRQVWUDLQW RI WKH ILUVW OLQH LV OHVV WKDQ RU HTXDO WR Notes
type which means the feasible solution zone lies towards the origin.
The no-shaded portion can be seen is the feasible area shown in Figure
1RWH ,I WKH FRQVWUDLQW W\SH LV WKHQ WKH VROXWLRQ ]RQH DUHD OLHV DZD\
from the origin in the opposite direction). Now the second constraints
line is drawn. When the second constraint is drawn, you may notice that
a portion of feasible area is cut. This indicates that while considering
both the constraints, the feasible region gets reduced further. Now any
point in the shaded portion will satisfy the constraint equations.
For example, let the solution point be (15, 20) which lies in the feasible region.
If the points re substituted in all the equations, it should satisfy the conditions.
2x1 + 3x2
2x1 + x2
Now, the objective is to maximize the profit. The point that lies at the
furthermost point of the feasible area will give the maximum profit. To
locate the point, we need to plot the objective function (profit) line.
Equate the objective function for any specific profit value Z,
Consider a Z-value of 60, i.e.,
6x1 + 4x2 = 60
Substituting x1 = 0, we get x2 = 15 and if x2 = 0, then x1 = 10
Therefore, the co-ordinates for the objective function line are (0, 15),
(10, 0) as indicated by dotted line L1 in Figure. The objective function
line contains all possible combinations of values of xl and x2.
The line L1 does not give the maximum profit because the furthermost
point of the feasible area lies above the line L1. Move the line (parallel
to line L1) away from the origin to locate the furthermost point. The
point P, is the furthermost point, since no area is seen further. Take the
corresponding values of x1 and x2 from point P, which are 15 and 30
respectively, and are the optimum feasible values of x1 and x2.
Therefore, we conclude that to maximize profit, 15 numbers of corru-
gated boxes and 30 numbers of carton boxes should be produced to get
a maximum profit.
Zmax = 6x1 + 4x2
= 6(15) + 4(30)
Maximum profit: Rs. 210.00
PAGE 9
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
1.6 Special Cases in Graphical Method
10 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
1.6.2 Multiple Solution
Illustration 3: Solve the following LPP using graphical method:
Maximize Z = 5x1 + 3x2
Subject to: 3x1 + 5x2 = 15
10x1 + 6x2 = 10
x 1, x 2 ! = 0
Solution: For the above problem it can be seen that second constraint is
the multiple of objective function. If we plot the second constraint and
the objective function then two lines will be parallel to each other. In
such situation, there are more than one optimal solution of the problem
though the value of objective function remains same.
PAGE 11
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
1.7 Simplex Method
The Graphical Method is applicable only to solve two variable linear
programming problem. For handling more than two variables we have
to use simplex method. However, a two-variable problem can also be
handled by simplex method.
Before proceeding with the algorithm of simplex method, we will discuss
some results that are being used to derive simplex method.
(i) If there is a feasible solution then there is a basic feasible solution
as well.
(ii) An optimal solution is at least one of the basic feasible solution.
(iii) If we have a basic feasible solution which is not optimal, then there
is a procedure to get a new basic feasible solution from it with an
improved value of objective function.
(iv) The improved basic feasible solution is
(a) Optimal, if (zj cj IRUDOOMIRUPD[LPLVDWLRQSUREOHPDQG
(zj cj IRU DOO M IRU PLQLPLVDWLRQ SUREOHP
(b) Unbounded, if for at least one j for which
yij L « P ]j cj) 0
§ m ·
where (zj cj ௗ ¨ ¦cBi yij c j ¸
©i1 ¹
Step 3: Convert all the inequations of the constraints into equations by Notes
introducing slack and/or surplus variables in the constraints. Put the costs
of these variables equal to zero.
Step 4: Obtain an initial basic feasible solution to the problem on the
form XB = B–1b
And put it in the first column of the simplex table.
¦
m
Step 5: Compute the net evaluation (zj cj) = c yij c j for all j =
i 1 Bi
1, 2, ... n
Examine the sign of zj cj
(i) If all (zj cj WKHQ WKH LQLWLDO EDVLF IHDVLEOH VROXWLRQ ;B is
an optimum basic feasible solution.
(ii) If at least one (zj cj) 0 proceed on to next step.
Step 6: If there are more than one negative (zj – cj), then choose the most
negative of them. Let it be zr – cr, for some j = r. Examine the following:
(i) If all yir i = 1, 2, … m), then there is an unbounded
solution to the given problem.
(ii) If at least one yir ! 0 (i = 1, 2, … m), then the corresponding
vector yr enters the basis yB.
Step 7: Compute the ratios
X Bi ½
® , yir ! 0, i 1, 2, ... m ¾ and select the minimum of them.
¯ yir ¿
X
Let the minimum of these ratios be Bk , then the vector yk will leave
ykr
the basis yB. The common element ykr is known as leading element or
pivotal element of the table.
Step 8: Convert the leading element to unity by dividing its row by the
leading element itself and all other element in its column to zeros by
using the relations:
y ykj
kj
, j 1, 2, ... n
ykr
ykj
And y ij yij yir , I = 1, 2, …. m, i k
ykr
PAGE 13
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes The above changes can be easily done using elementary row operations:
Cj 10 15 20 0 0
Basic
CBi Variables X1 X2 X3 S1 S2 Solution B Ratios
0 S1 2 4 6* 1 0 24 24/4
0 S2 3 6 0 1 30 30/6
Zj – Cj –10 –15 – 20 0 0 0
Since all the values of Zj – Cj are not greater than equal to zero. Hence
the initial solution is not optimal. The most negative Zj – Cj corresponds
to X3. Therefore the column vector X3 enters the basis. Further the mini-
mum ratio is 4 which corresponds to the variable S1. Thus, S1 will leave
the basis and the element y13 is the leading element for the first iteration.
14 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes The algorithm will remain same as simplex method except at the last iteration
of simplex method there can arise any one of the following three cases:
(i) There is no vector corresponding to some artificial variable, in the
basis yB. In such a case we continue simplex method until optimality
reached.
(ii) There is at least one vector corresponding to artificial variable in
the basis with zero value. Also the coefficient of M in each net
evaluation Zj – Cj is nonnegative. In such a case the current basic
feasible solution is a degenerate. In this case an optimum solution
to the given LPP includes an artificial variable and an optimum
basic feasible solution does exist.
(iii) At least one artificial variable is present in the basis with positive
value and the net evaluation are all non-negative. In this case, the
LPP does not possess any feasible solution.
Note that while applying simplex method, whenever a vector corresponding
to some artificial variable leave the basis, we drop the column of that
vector from simplex table.
Illustration 6: Solve the following LPP using Big M method.
Minimize Z = 2x1 + x2
Subject to 3x1 + x2 = 3
4x1 + 3x2 t 6
x1 + 2x2 d 3
x 1, x 2 ! 0
Solution: Since the problem is of minimising the objective function, we
first convert it into that of maximisation, using the relation.
Min Z = –Max (–Z) = –2x1 – x2
Introducing Artificial variables xA1 ! 0, xA2 ! 0 and surplus variable S1 ! 0,
slack variable S2 ! 0. Thus the given LPP can be rewritten as
Maximize Z 2 x1 x2 0 S1 0 S 2 Mx A1 Mx A 2
Subject to 3 x1 x2 x A1 3
4 x1 3 x2 S1 x A 2 6
x1 2 x2 S 2 3
x1 , x2 , S1, S 2, x A1 , x A 2 t 0
16 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
An initial basic feasible solution to the given problem is xA1 = 3 and xA1 = 6, Notes
S2 = ± 0
The initial Simplex table for the problem will be
Cj –2 –1 –M 0 –M 0
Basic Solution
CBi Variables X1 X2 XA1 S1 XA2 S2 B Ratios
–M XA1 3* 1 1 0 0 0 3 3/3
–M XA2 4 3 0 –1 1 0 6 6/4
0 S2 1 2 0 0 0 1 3 3/1
Zj – Cj –7M + 2 –4M + 1 0 M 0 0 ±0
Notes
1.7.3 Two-Phase Simplex Method
Two-Phase simplex method is an alternative procedure to solve LPP
involving artificial variables. It applies simplex algorithm in two phases.
In Phase I, the simplex algorithm tests the feasibility (consistency) of
the problem. If the LPP is feasible then it determines the starting basic
feasible solution. Phase II then leads from the starting basic feasible
solution determined at the end of Phase I and uses simplex algorithm to
arrive at an optimal feasible solution, if any.
Algorithm for Two-Phase method:
Phase I
Step 1: Form a modified problem for phase I from the canonical form
of the original problem by replacing the objective function with the sum
of only the artificial variables along with the same set of constraints of
the canonical form of the original problem.
Minimize Z * A1 A2 }}} An
18 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
1.7.4 Identification of Special Cases from Simplex Table
Infeasible Solution: In the optimal simplex table, if the set of basic
variables contains at least one artificial variable, then the given problem
is said to have infeasible solution.
Unbounded Solution: If the constraints coefficients of the entering
variable in this simplex table are either less than or equal to 0 then the
solution space is unbounded and has no finite optimum solution.
Alternate Optimum Solution: In the optimum simplex table if the net
evaluation Zj Cj value for at least one non-basic variable is equal to 0,
then the problem is said to have alternate optimum solution.
1.8 Duality
The term duality implies that every linear programming problem whether
of maximization or minimization has related with it another linear pro-
gramming problem based on the same data. The original (given) problem
is called the primal problem while the other i.e. the converted form, is
called its dual problem. In general, either problem can be considered as
primal and the other as its dual. Thus, the two problems constitute the
pair of dual problems.
Notes 5. The number of variables in the dual problem is equal to the number
of constraints in the primal problem.
6. The number of constraints in the dual problem is equal to the
number of variables in the primal problem.
7. If ‘C’ is the cost vector of the primal problem then its transpose
becomes the resource vector for the dual problem.
8. If ‘B’ is the resource of the primal problem then its transpose
becomes the cost vector for the dual problem.
9. If ‘A’ is the coefficient matrix in the constraints of primal problem
then its transpose becomes the coefficient matrix for constraints in
the dual problem.
Illustration 7: Form the dual of the following primal problem.
Maximize Z = 4x1+ 10x2 + 20x3
Subject to: 2x1 + 4x2 + 8x3
4x1 x2 + 8x3
6x1 + 8x2 + 2x3
x1, x2 and x3
Solution:
The given problem is termed as a primal problem. Let Yi be dual variable
associated with the ith constraint of the primal problem. Let’s write the
Coefficient matrix A, Cost Vector C and resource vector B.
ª2 4 8º ª 25º
A «« 4 9 8 »» , B «30 » , C
« » >4 10 20@
«¬ 6 8 2 »¼ «¬ 40 »¼
20 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
1.8.2 Characteristics of Duality
1. Dual of dual is primal.
2. If either the primal or dual problem has a solution, then the other
also has a solution and their optimal values are equal
3. If any of the two problems has an infeasible solution, then the value
of the objective function of the other is unbounded.
4. The value of the objective function for any feasible solution of
the primal is less than the value of the objective function for any
feasible solution of the dual.
5. If either the primal or dual has an unbounded solution, then the
solution to the other problem is infeasible.
6. If the primal has a feasible solution, but the dual does not then the
primal will not have a finite optimal solution and vice versa.
PAGE 21
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
° ( z j c j ) °½ ° ( z j c j ) °½
Max ® ¾ d 'Ck d Min ® ¾
°̄ Ykj ! 0 °¿ °̄ Ykj 0 °¿
Cj 10 15 20 0 0
CBi Basic Variables X1 X2 X3 S1 S2 Solution B
20 X3 0 –1 1 1/2 –1/3 2
10 X1 1 5 0 –1 1 6
Zj – Cj 0 15 0 0 10/3 100
22 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Here, C1 and C3 are the cost coefficient of basic variable and C2 is Notes
associated with non-basic variable in the optimal table.
Range for C2 ¨F2 z2 – c2 ¨ c2
15 10 / 3 ½
Range for C1: Max ® , ¾ d 'ck ± d 'ck
¯ 5 1 ¿
15 10 / 3 ½
Range for C3: 'ck d Min ® , ¾ c1
¯ 1 1 / 3 ¿
Notes §1 / 2 1 / 3 ·
B-1 = (S1 S2) = ¨ ¸ Also XB = [2, 6]
© 1 1 ¹
2 ½ 6 ½
Range for b1: Max ® ¾ d 'b1 d Min ® ¾ = !–4 d 'b1 d 6
¯1 / 2 ¿ ¯ 1 ¿
6 ½ 6 ½
Range for b2: Max ® ¾ d 'b2 d Min ® ¾ = !–6 d 'b2 d 18
¯1 ¿ ¯ 1 / 3 ¿
IN-TEXT QUESTIONS
1. What is the primary focus of linear programming in management?
(a) Maximizing profits
(b) Optimizing resources allocation
(c) Minimizing risks
(d) Enhancing customer satisfaction
2. In linear programming, what are the assumptions typically made
about the decision-making environment?
(a) Perfect competition
(b) Unlimited resources
(c) Certainty in parameters
(d) No constraints
3. What is the purpose of the graphical method in linear programming?
(a) To visualize data
(b) To identify optimal solutions
(c) To solve nonlinear problems
(d) To analyze financial statements
4. Which method is commonly used to solve large-scale linear
programming problems iteratively?
(a) Graphical method
(b) Simplex method
(c) Sensitivity analysis
(d) Duality method
24 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 25
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
1.10 Summary
The basic theme of the lesson is to make learner understand the concept
of linear programming problem and the tools to get the solution of the
same. To solve an LPP for two variables one may use graphical method
but for more than two variables simplex method has to be used. For
special type of constraints, Big M method and two-phase method may
be used. In the later part of the lesson concept of duality and sensitivity
analysis has been introduced.
26 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
1.12 Self-Assessment Questions
1. A firm manufacturing two types of electrical items A and B, can
make a profit of Rs. 120 per unit of A and Rs. 200 per unit of B.
Both A and B make use of two necessary components, a resistor
and a capacitor. Each unit of A requires 3 resistors and 2 capacitors
and each unit of B requires 2 resistors and 4 capacitors. The total
supply of components per month is restricted to 210 resistors and
300 capacitors. Type B is an export model requiring a transistor,
which has a supply restricted to 65 units per month. How many each
of A and B should the firm manufacture per month to maximize its
profit? How much is this profit?
2. A company produces two types of belts. Belt X is of good quality
and belt Y is of a lower quality. Profits on the two types of belts are
Rs. 4 and Rs. 3 per belt respectively. Each belt of type X requires
twice as much time as required by a belt of type Y. If all belts are
of Type Y, then the company could produce 1000 belts per day. But
the supply of raw material is sufficient only for 800 belts per day.
Belt X requires a fancy buckle and only 400 buckles are available
per day. For belt of Type Y only 700 buckles are available per
day. How should the company manufacture the two types of belts
in order to have a maximum overall profit?
3. A publisher sells a hard cover edition of a text book for Rs. 72
and a paperback edition of the same text for Rs. 40. Costs to the
publisher are Rs. 56 and Rs. 28 per book respectively in addition
WRZHHNO\FRVWVRI5V%RWKW\SHVRIERRNVUHTXLUHPLQXWHV
of printing time, although hard cover requires 10 minutes binding
time and the paperback requires only 2 minutes. Both the binding
and printing operations have 4800 minutes available each week.
How many of each type of book should be produced in order to
maximize profit?
4. List and explain the assumptions of linear programming problem.
5. A small manufacturer employs 5 skilled men and 10 semi-skilled
men for making a product in two qualities: a deluxe model and an
ordinary model. The production of a deluxe model requires 2-hour
work by a skilled man and 4-hour work by semi-skilled man. The
PAGE 27
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes ordinary model requires 1-hour work by a skilled man and 3-hour
work by semi-skilled man. According to worker union’s rule, no
man can work more than 8 hours per day. The profit of the deluxe
model is Rs. 1000 per unit and that of the ordinary model is Rs. 800
per unit. Formulate a linear programming model for this problem
to determine the production volume of each model such that profit
is maximised.
6. Solve the following LPP:
Maximize Z = x1 + 4x2
Subject to: 2x1 + 4x2
5x1+3x2
x1, x2
7. Use graphical method to solve the following LPP:
Maximise Z = 2x1 + x2
Subject to x1 + 2x2
x1 + x2
x1 – x2
x1 – 2x2
x1, x2
8. Use Big M method to solve the following LPP
Minimize Z = 2x1 + 3x2
Subject to x1 + x2 = 6
7x1 + x2
x1, x2
)LQG DQ\ WKUHH DOWHUQDWH RSWLPDO VROXWLRQ LI WKH\ H[LVW IRU WKH
following LPP:
Maximize Z = 2x1 + 4x2
Subject to x1 + 2x2
x1 + x2
x1, x2
28 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Basic x1 x2 x3 x4 x5 x6 x7 Solution
Z 1 0 3 0 0 2 M 8
x4 3 0 2 1 0 1 0 6
x2 0 1 1 0 0 1 0 4
x5 1 0 3 0 1 4 1 10
PAGE 29
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
1.13 References
Hamdy A. Taha. Operation Research An Introduction, Pearson
Publication, 10th Edition.
Panneerselvam R. Operations Research, PHI learning, Second edition.
30 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
2
Application of LP
Marketing, Finance,
Operations Management
Dr. Abhishek Kumar Singh
Assistant Professor
PGDAV(E) College
University of Delhi
Email-Id: abhishekbhu008@gmail.com
STRUCTURE
2.1 Learning Objectives
2.2 Linear Programming
2.3 Application of LP in Marketing
2.4 Application of LP in Finance
2.5 Application of LP in Operations Management
2.6 Summary
2.7 Answers to In-Text Questions
2.8 Self-Assessment Questions
2.9 References and Suggested Readings
PAGE 31
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
2.2 Linear Programming
Linear programming is a mathematical technique for allocating scarce
resources. The technique is frequently used to maximise the value of a
goal, such as minimising costs or increasing profits while dealing with
constraints. In layman’s terms, this means that the technique is intended
to assist marketers or managers in planning and making better resource
allocation decisions. In fact, linear programmes enable marketers to make
the best use of limited resources, such as money, manpower, materials,
time, or skills, to name a few.
Marketers must convert the given problem into a linear programming
model, using the appropriate variables and standard techniques to solve
the real-world problem. The model includes the following components:
1. Variables representing the available options.
2. The mathematical expression relating to the variables and conditions
at hand.
According to Bazaraa, Jarvis, and Sherali (2011), linear programming is
a mathematical model with a goal function and constraints that can be
used to help a business decide how to best utilise its current market and
generate sustainable profits.
Linear programming generates solutions based on the characteristics and
characteristics of the actual situation or problem. As a result, the appli-
cation of linear programming is very broad, and it can find solutions in
a variety of fields such as marketing. In fact, this technique has proven
to be extremely effective in marketing and distribution management,
among other areas. A marketing manager, for example, can use linear
programming to distribute a fixed advertising budget across multiple
media channels in order to maximise audience exposure.
Thousands of decision variables and constraints are used in the model
to solve practical problems in marketing research. The underpinning
strength is the use of the extreme point of a function to satisfy certain
constraints (Bazaraa et al., 2011). In this situation, the constraints on the
problem are linear, and the function in the linear equation that is used
to optimise solutions for the marketing research problem is known as
the objective function.
32 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 33
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
34 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 35
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
36 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
2.2.7 Assumptions
Linear Programming is based on the following assumptions:
1. Purpose All of the relationships between function and constraints are
linear.
This assumption has the following corollaries:
(a) We assume no economies of scale or diseconomies of scale; six
units of product T require six times the amount of raw material as
one unit of T.
(b) We assume that the decision variables do not interact; the total raw
material required for X and T is simply the sum of the individual
requirements for X and K 2.
There is only one goal function. There was only one goal in our product-
mix problem, and that was to maximise profit. However, it is not easy in
practise. An organisation may have a number of objectives for a specific
decision, with possibly some priorities in between, all of which must be
considered together.
PAGE 37
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes depending on the needs and prerequisites of the target market can be
advantageous for the organisation if management optimises the linear
programming model.
For instance, to carry out a marketing/consumer research programme, a
company might decide to use a variety of techniques to reach the target
audience and decide on the best combination of product characteristics to
address the needs and desires of the customer using the suitable media
outlet (Zopounidis and Pardalos, 2010). The available budget, restrictions,
and total amount of money are what determine which media options are
available. Media outlets include, but are not limited to, magazines, radio,
social media, and other platforms that can be tailored to reach the target
audience. The linear programming method can also be used to choose the
best advertising budget from a variety of advertising channels.
A common example is the use of linear programming to determine the
effectiveness of available advertising alternatives using the formula:
WAA = Adjusted audience (AA) × Exposure factor (E × F) × Evaluation
Factor (E v F)
In the above formula, AA is the target audience, E × F is the fraction
of the audience that reads the advertisement, and (E v F) is a subjective
weight that is based on a scale of 0 1 score. Assume that 4 advertising
vehicles have been considered, the objective function is expressed as:
19
To determine the best advertising vehicle, enter the formula into Excel
and run an excel solver.
Example: The Dyson Company, which recently launched a new line
of hair appliances, has $282,000 to spend on advertising. The product
will first be tested in the Dallas area. The funds will be used for a TV
advertising blitz over one weekend in November (Friday, Saturday, and
Sunday). Daytime advertising, evening news advertising, and Sunday
game-time advertising are the three options available. A mix of one-
minute TV spots is preferred.
38 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
40 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 41
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
42 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 43
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Utilizing resources effectively, that is, getting the most out of them or Notes
reducing their loss, underutilization, or waste, is the main focus of oper-
ations management. The percentage of time that is used or occupied, the
amount of space used, the intensity of activity, etc. can all be used as
indicators of how fully the resources are being used. Each metric shows
how much of these resources’ potential or capacity is actually being used.
This is referred to as the resource utilisation objective.
The achievement of both satisfactory customer service and resource util-
isation is a concern of operations management. A change in one will fre-
quently result in a decline in the other. Often, neither can be maximised,
so both must be accomplished with a satisfactory level of performance.
These two goals must be the focus of all operations management activ-
ities, and operations managers will encounter many issues as a result of
this conflict. Therefore, operations managers need to try to balance these
fundamental goals.
PAGE 45
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
46 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Method Input per run (units) Output per run (units) Notes
Raw Raw
material 1 material 2 Part A Part B
1 7 5 6 4
2 4 7 5 8
3 2 7 3
6 X1 5 X 2 7 X 3 4 X1 8 X 2 3 X 3
and
5 4
Z = Minimum of §¨ 6 X 1 5 X 2 7 X 3 and 4 X 1 8 X 2 3 X 3 ·¸
© 5 4 ¹
Constraints are on the availability of raw materials. They are, for
PAGE 47
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes proportional). However, the above model can be easily traduced to the
generally acceptable linear programming format.
§ 6 X1 5 X 2 7 X 3 4 X1 8 X 2 3 X 3 ·
Let, y = Min ¨ and ¸
© 5 4 ¹
6 X1 5 X 2 7 X 3 4 X1 8 X 2 3 X 3
It follows that \ DQG \
5 4
i.e., 6 X 1 5 X 2 7 X 3 t 0 and 4 X 1 8 X 2 3 X 3 t 0
5 X 1 7 X 2 9 X 3 d 240
6 X1 5 X 2 7 X 3 t 0
4 X1 8 X 2 3 X 3 t 0
48 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 49
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
50 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
52 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
10. What is the primary purpose of the practice questions section? Notes
2.6 Summary
Marketing Application
Media Selection
Media selection is one application of linear programming in marketing.
LP can assist marketing managers in allocating a set budget to
various advertising media.
The goal is to maximise exposure reach, frequency, and quality.
Restrictions on allowable allocations typically arise as a result of
company policy, contract requirements, and media availability.
Marketing Research
Marketing research is conducted by a company to learn about
consumer characteristics, attitudes, and preferences.
Marketing research services include creating the study, conducting
surveys, analysing the data gathered, and making recommendations
to the client.
Targets or quotas for the number and types of respondents to be
surveyed may be established during the research design phase.
The marketing research firm’s goal is to conduct the survey in
such a way that it meets the client’s needs at the lowest possible
cost.
Financial Application
LP can be used in capital budgeting, make-or-buy, asset allocation,
portfolio selection, financial planning, and other financial decisions.
Portfolio selection problems entail selecting specific investments,
such as stocks and bonds, from a pool of investment options.
PAGE 53
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
54 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes stock and 4,000 m3 of non-refrigerated stock. The cost per kilometre
of a Type A is $30 and $40 for Type B. How many trucks of each
type should the grocer rent to achieve the minimum total cost?
$ VFKRRO LV SUHSDULQJ D WULS IRU VWXGHQWV 7KH FRPSDQ\ WKDW LV
providing the transportation has 10 buses of 50 seats each and 8 buses
RIVHDWVEXWRQO\KDVGULYHUVDYDLODEOH7KHUHQWDOFRVWIRUDODUJH
bus is $800 and $600 for the small bus. Calculate how many buses
of each type should be used for the trip for the least possible cost.
10. A store wants to liquidate 200 of its shirts and 100 pairs of pants
from last season. They have decided to put together two offers, A
and B. Offer A is a package of one shirt and a pair of pants which
will sell for $30. Offer B is a package of three shirts and a pair
of pants, which will sell for $50. The store does not want to sell
less than 20 packages of offer A and less than 10 of offer B. How
many packages of each do they have to sell to maximize the money
generated from the promotion?
3
Transportation Problem
Dr. Abhishek Kumar Singh
Assistant Professor
PGDAV(E) College
University of Delhi
Email-Id: abhishekbhu008@gmail.com
STRUCTURE
3.1 Learning Objectives
3.2 Introduction
3.3 Transportation Problem
3.4 Some Special Codes of TP
3.5 Summary
3.6 Answers to In-Text Questions
3.7 Self-Assessment Questions
3.8 References
3.9 Suggested Readings
PAGE 57
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
3.2 Introduction
This chapter facilitates an overview of the notion about transhipment
issue is a special Linear Programming Problem (LLP) because it takes
into account the presumption that all sources and sinks can simultane-
ously accept and distribute shipments (function in both directions). The
goal of the transportation model, a special form of linear programming
issues, is to reduce the overall cost of moving commodities from the
various supply origins to the various demand destinations and the basic
goal of an assignment problem is to reduce the overall amount of time
needed to accomplish a group of activities, or to increase skill ratings,
or to reduce the cost of the assignments.
58 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 59
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes The first m constraints, which are known as the supply limitations, state
that the quantity of commodity units that are readily available at each
origin must not be exceeded. The next n limitations guarantee that the
requirements for commodity units at destinations will be met. Since they
indicate the quantity of transferred commodity units, the decision variables
are characterised as positive.
Here is a common form of the transportation issue.
60 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
Destination
D1 D2 D3 D4 Supply
Q1 3 1 7 4 300
Q2 2 6 5 9 400
Source
Q3 8 3 3 2 500
Demand 250 350 400 200 1200
Explanation: Given three origins (O1, O2, and O3) and four destinations
(D1, D2, D3, and D4) For the sources O1, O2, and O3, the supplies are
300, 400, and 500, respectively. Demands are respectively 250, 350, 400,
and 200 for the destinations D1, D2, D3, and D4.
Solution: According to the Northwest Corner approach, the beginning
point must be the north-western corner of the table, or (O1, D1). When
determining the price per transportation, every single value in a cell is
taken into consideration. After evaluating the supply from source O1
and demand for column D1 in the figure, a minimum of two should be
awarded to the cell (O1, D1). Since there is no longer any requirement
for Column D1, the entire Column D1 will be cancelled. The quantity
coming from source O1 is still 300 – 250 = 50.
PAGE 61
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Destination
D1 D2 D3 D4 Supply
250
Q1 3 1 7 4 300 50
Q2 2 6 5 9 400
Source
Q3 8 3 3 2 500
250
Demand 0 350 400 200 1200
Examine the northwest quadrant of the table, which is (O1, D2), but leave
off column D1. Then, distribute the bare minimum of supply among the
corresponding column and rows. The source O1 should only allocate 50 to
the cell, which is less than what is required for D2 (i.e., 350). (O1, D2).
Cancel row O1 because the supply from row O1 is finished. Column D2
still has a demand of 350 – 50 = 50.
Destination
D1 D2 D3 D4 Supply
250 50
Q1 3 7 4 30050 0
Q2 2 6 5 9 400
Source
Q3 8 3 3 2 500
250 350
Demand 0 300 400 200 1200
Finding the Northwest corner of the remaining table, O2, D3, now compare
the O2 supply (100) to the D2 demand (400) and assign the smaller (100)
to the cell (O2, D2). Cancel the row O2 since the supply from O2 has been
fully satisfied. 400 – 100 = 300 is the remaining demand for column D3.
Destination
D1 D2 D3 D4 Supply
250 50
Q1 3 7 4 30050 0
300 100
Q2 2 6 5 9 400 100 0
Source
Q3 8 3 3 2 500
350
250 300 400
Demand 0 0 300 200 1200
62 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Proceeding in the same way, the final values of the cells will be: Notes
Destination
D1 D2 D3 D4 Supply
250 50 100
Q1 3 7 4 30050 0
300 300 300
Q2 2 6 5 9 400 100 0
Source
Q3 8 3 3 2 500 200 0
350 400
250 300 300 200
Demand 0 0 0 0 1200
Note that in the last remaining cell, which was cell, the demand for the
pertinent columns and rows is equal (O3, D4). Since O3 was in supply
and D4 was in demand, the cell in this case received 200. The last item
was that neither a row nor a column contained anything.
Simply multiply the value of each cell by the designated value to obtain
the basic solution, which is (250 * 3) + (50 * 1) + (300 * 6) + (100 * 5) +
(300 * 3) + (200 * 2) = 4400.
Destination
D1 D2 D3 D4 Supply
Q1 3 1 7 4 300
Q2 2 6 5 9 400
Source
Q3 8 3 3 2 500
Demand 250 350 400 200 1200
The Least Cost Cell approach requires that the least expensive cell in
the table be identified, and that cell is 1 (cell (O1, D2)).
Now contrast the supply in row O1 with the demand in column D2 and
assign the cell with the lower number. Give the cell 300 because it is the
lesser value. As the supply from O1 is exhausted, the demand for column
D2 is 350 – 300 = 50, hence this row should be cancelled.
PAGE 63
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Destination
D1 D2 D3 D4 Supply
Q1 3 300 1 7 4 300 0
Source Q2 2 6 5 9 400
Q3 8 3 3 2 500
350
Demand 250 50 400 200 1200
Find the cell with the lowest cost among the remaining cells next. Cells
(O2, D1) and (O3, D4) are the two least expensive, both costing 2. Let’s
decide (O2, D1). Next, ascertain the supply and demand for the relevant
cell. Then, allocate the smallest sum between the two to the cell. Finally,
remove the row or column whose supply or demand is now zero as a
result of the allocation
Destination
D1 D2 D3 D4 Supply
300
Q1 3 1 7 4 300 0
2 400
Source 250
Q2 6 5 9 150
Q3 8 3 3 2 500
250 350
Demand 0 50 400 200 1200
Now, (O3, D4), with a cost of 2, is the cell with the lowest cost. As the
supply is greater than the demand, assign 200 to this cell. Therefore, the
column is removed.
Destination
D1 D2 D3 D4 Supply
300 1
Q1 3 7 4 300 0
250 9 400
200
Source Q2 2 6 5 150
500
Q3 8 3 3 2 300
250 350 200
Demand 0 50 400 0 1200
64 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
The two unallocated cells with the lowest costs are the other two. Please Notes
choose any at random (O3, D2). From the supply in the corresponding row
and the demand in the corresponding column, enter the smallest amount
you can into this cell. Remove the row or column with a value of zero.
Destination
D1 D2 D3 D4 Supply
300 1
Q1 3 7 4 300 0
250
Q2 2 6 5 9 400 150
Source 50
200
Q3 8 3 3 2 500300 250
350
250 50 200
Demand 0 0 400 0 1200
Currently, the cheapest cell is (O3, D3). Eliminate the row or column
with a value of zero and distribute the minimum of supply and demand
Destination
D1 D2 D3 D4 Supply
300
Q1 3 1 7 4 300 0
250
Q2 2 6 5 9 400 150
Source 50 250
200
Q3 8 3 3 2 500300250 0
350
250 50 400 200
Demand 0 0 150 0 1200
There is only one cell left, (O2, D3), and it costs $5. Its supply is 150
and demand is 150, hence demand is 150 and both are equal in quantity.
Allocate it to this cell.
PAGE 65
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Destination
D1 D2 D3 D4 Supply
300 1
Q1 3 7 4 300 0
250 150
200
9
Q2 2 6 250 5 400150 0
50
200
Q3 8 3 3 2 500300250 0
350 400
250 50 150 200
Demand 0 0 0 0 1200
After computation of the cost by least cost method, simply multiply the
cost of the cell by the values assigned to it, then add the results to get
the fundamental solution, which is (300 * 1) + (25 * 2) + (150 * 5) +
(50 * 3) + (250 * 3) + (200 * 2) = 2400
Destination
D1 D2 D3 D4 Supply
Q1 3 1 7 4 300
Q2 2 6 5 9 400
Source
Q3 8 3 3 2 500
Demand 250 350 400 200 1200
The answer is to first identify the lowest value for each row, then the
next lowest. The corresponding row difference should then be filled in
with the absolute difference between these two lowest values, as seen in
the image below. The second-least value, 3, and the two least values in
row O1 are separated by an absolute difference of 2. Rows O2 and O3
have identical absolute differences of 3 and 1, respectively.
The lowest value should be found for each column, then the next lowest.
The corresponding column difference should then be filled in with the
absolute difference between these two lowest values, as shown in the
66 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
image. Column D1’s least value is 2, followed by its second-least value, Notes
3, and their absolute difference is 1. Columns D2, D3, and D3 all have
absolute differences of 2, 2, and 2, respectively
Destination
Row
D1 D2 D3 D4 Supply difference
Q1 3 1 7 4 300 2
Source Q2 2 6 5 9 400 3
Q3 8 3 3 2 500 1
Demand 250 350 400 200 1200
Column
Difference 1 2 2 2
These numbers for the row and column disparity are also known as
penalties. Pick the most severe penalty right now. Three strikes, or row
O2, is the maximum penalty. Then, by choosing the cell in row O2 with
the lowest cost, divide the minimal amount between the supply of the
corresponding row and the demand of the corresponding column. Since
demand is lower than supply, put 250 in the cell to reflect the desire for
the column. Delete column D1 next.
Destination
Row
D1 D2 D3 D4 Supply difference
Q1 3 1 7 4 300 2
250 400
Source Q2 2 6 5 9 150 3
Q3 8 3 3 2 500 1
250
Demand 0 350 400 200 1200
Column
Difference 1 2 2 2
PAGE 67
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes From the remaining cells, find out the row difference and column
difference.
Row
D1 D2 D3 D4 Supply difference
Q1 3 1 7 4 300 2 3
250 400
Source Q2 2 6 5 9 150 3 1
Q3 8 3 3 2 500 1 1
250
Demand 0 350 400 200 1200
Column
Difference 1 2 2 2
- 2 2 2
Row
D1 D2 D3 D4 Supply difference
300
Q1 3 1 7 4 300 0 2 3
250 400
Source Q2 2 6 5 9 150 3 1
Q3 8 3 3 2 500 1 1
250 350
Demand 0 50 400 200 1200
Column
Difference 1 2 2 2
- 2 2 2
68 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Now find the row difference and column difference from the remaining cells. Notes
Destination
Row
D1 D2 D3 D4 Supply
difference
300
Q1 3 1 7 4 300 0 2 3
-
250 400
Q2 2 6 5 9 3 1 1
150
Source
Q3 8 3 3 2 500 1 1 1
250 350
Demand 400 200 1200
0 50
Column
1 2 2 2
Difference
- 2 2 2
- 3 2 7
You should now select the maximum penalty, which is 7 and falls under
column D4. The cheapest cell in column D4 is (O3, D4), which costs 2.
More cell phones are available than are needed (O3, D4).
Remove the column and give the cell $200.
Destination
Row
D1 D2 D3 D4 Supply
difference
300
Q1 3 1 7 4 300 0 2 3
-
250
Q2 2 6 5 9 400 150 3 1 1
Source 200
Q3 8 3 3 2 500 300 1 1 1
PAGE 69
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Find the row difference and the column difference from the remaining cells.
Destination
Row
D1 D2 D3 D4 Supply
difference
300
Q1 3 1 7 4 300 0 2 3
-
250
Q2 2 6 5 9 400 150 3 1 1
Source 200
Q3 8 3 3 2 500 300 1 1 1
70 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Now there is only one column so select the cell with the least cost Notes
and allocate the value.
PAGE 71
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes There is no equilibrium left. Therefore, multiply each cell’s assigned value
by its associated cell cost, then sum them all up to get the total cost,
which is (300 * 1) + (250 * 2) + (50 * 3) + (250 * 3) + (200 * 2) +
(150 * 5) = 2850.
72 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
As a result, there are more loops overall, which makes calculating more Notes
difficult and time-consuming. Also, in the MoDi technique, only one
closed path is traced for the most negative cell, and cell evaluations
of all vacant cells are calculated simultaneously. In comparison to the
Stepping Stone Method, it thus saves time.
The MoDi method is often referred to as the “u v Method.”
Let us now examine how the Modified Distribution Method (MoDi)
functions. The following numerical explanation can help you grasp it:
Applying Modified Distribution Method (MoDi) Numerical
Question: Verify the optimality of the fundamentally workable solution
provided below using the Modified Distribution Method (MODI) approach.
Solution: Take note that “CYAN” is the only colour used throughout the
discussion. If there are any steps in the exam, you must only write the
items that are shown in this regular colour underneath them; otherwise,
you can simply use this matrix to solve the issue as it is presented here.
ĺ &UHDWH D FRVW PDWUL[ E\ HQWHULQJ FRVW LQIRUPDWLRQ LQ GHVLJQDWHG FHOOV
ĺ ,QWURGXFH GXDO YDULDEOHV X Y WKDW UHSUHVHQW WKH OLPLWDWLRQV RQ
supply and demand now.
Note, the dual variables ui and vj in this case are set up so that ui + vj = cij
(where i = 1, 2, 3,... m & j = 1, 2, 3,... n).
PAGE 73
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
Let us now conduct a Vacant Cell Assessment. With each empty call, a
difference is taken in order to evaluate vacant cells.
?
Analysis of cell = Cij(ui + vj), where Cij = Value of unoccupied cell
Recall that the term “implicit cost” is “(ui + vj)”.
Implicit cost is the total of the dual variables for the row and columns
for each vacant cell.
The following formula should be used to fill in the empty cells:
(ii) Zero (0) – In this case, the fundamental feasible solution (BFS) Notes
is also the best possible option, but different allocations are still
viable at the same transportation cost.
(iii) Negative (-) – If this is true, the obtained solution is not ideal and
room for improvement still exists.
The current answer is the best one since all of the values in the unallo-
cated cells have now been determined to be positive (+) by employing
vacant cell evaluation.
The Stepping Stone Method can also be used to determine the answer
to this numerical problem.
PAGE 75
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
The basic feasible solution by (LCM) is: x11 = 150, x12 = 130, x13 = 30,
x23
x24 = 10 and x34
Once the basic feasible solution is found, next we have to determine the
optimum solution. The MoDi (Modified Distribution Method) approach
is used in this case. With this approach, we are able to obtain
u1+ v1 X1 + v2 X1 + v3 X2 + v3 = 4, u2 + v4 = 3, u3 + v4 = 0.
Taking u1 = 0 arbitrarily we obtain
u1 = 0, v1 = 2, v2 DQGY3 = 2, u2 = 2, v4 = 1, u3 í
76 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
3.4.2 Degeneracy
When there is less than m + n 1 positive Xij, or occupied cells, in a
basic feasible solution to a transportation problem with m origins and n
destinations, the problem is said to be a degenerate transportation problem.
Although the degeneracy problem may not provide a major challenge, it
can complicate computations when seeking the best minimum solution.
As a result, it is critical to spot degenerate problems early on and take
the appropriate steps to prevent any computational difficulty. The results
show that a degenerate basic viable solution only exists in transportation
problems when some partial sum of supply (row) equals some partial
sum of demand (column).
For example, the transportation issue below is degenerate. Due to the fact
WKDW LQ WKLV LVVXH D E DQG D D E E
PAGE 77
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes The degenerate problems can be resolved using a method called pertur-
bation.
Perturbation Method:
If we make sure that no partial sum of ai (supply) and bj (demand) is
equal, the degeneracy of the transportation problem can be avoided.
We build up a new problem with the following variables: ai = ai + d,
i = 1, 2,......, m bj = bj, j = 1, 2,......, and n 1, bn = bn + md, d ! 0.
No partial sum of ai can equal bj in this modified problem because of
the way it is built.
When the issue is resolved, we replace d = 0 to achieve the best reso-
lution of the initial issue.
Example:
78 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Then, in order to solve, you must balance supply and demand. As no Notes
shipment is really made when a dummy source and dummy destination
are used, the unit transportation cost for the dummy column and dummy
row are given zero values.
PAGE 79
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
IN-TEXT QUESTIONS
1. In a transportation problem where the demand or requirement is
equal to the available resource is known as
(a) Balanced transportation problem
(b) Regular transportation problem
(c) Resource allocation transportation problem
(d) Simple transportation model
2. An alternative optimal solution to a minimization transportation
problem exists whenever opportunity cost corresponding to
unused route of transportation is:
(a) positive and greater than zero
(b) positive with at least one equal to zero
(c) negative with at least one equal to zero
(d) none of the above
3. One disadvantage of using the North-West Corner Rule to find
initial the solution to the transportation problem is that
(a) it is complicated to use
(b) it does not consider the cost of transportation
(c) it leads to a degenerate initial solution
(d) all of the above
4. When total supply is equal to total demand in a transportation
problem, the problem is said to be:
(a) balanced
(b) unbalanced
(c) degenerate
(d) none of these
80 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 81
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
14. Which section of the lesson typically provides a brief explanation Notes
of specialized terms used in transportation problems?
(a) Learning Objectives
(b) Glossary
(c) Self-Assessment Questions
(d) Suggested Readings
15. What is the main purpose of the self-assessment questions
section?
(a) To introduce advanced topics
(b) To summarize key concepts
(c) To reinforce learning through practice
(d) To provide additional reading material
16. In transportation problems, what does an unbalanced situation
refer to?
(a) When supply exceeds demand
(b) When demand exceeds supply
(c) When transportation costs are high
(d) When transportation routes are unavailable
17. Which method in transportation problem solving is known for
its ability to produce more accurate results but may require
more computational effort?
(a) Northwest Corner Method
(b) Least Cost Cell Method
(c) Vogel’s Approximation Method
(d) Modified Distribution Method
3.5 Summary
Lesson 3 introduced the concept of the Transportation Problem, which
is a classic problem in operations management involving the efficient
allocation of goods from multiple sources to multiple destinations while
PAGE 83
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
84 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
3.7 Self-Assessment Questions
(I) Short Answer Questions:
1. What is the transportation problem, and how can it be resolved?
2. Describe the transportation model.
3. What is the most effective way to tackle TP?
4. What method of transportation is used?
5. What distinguishes transhipment from transit?
(II) Long Answer Questions:
1. Compare and contrast the stepping stone method with the MoDi
method for resolving a transportation issue.
2. A factory has four stores, S1, S2, S3, and S4, which are supplied by
three warehouses, W1, W2, and W3. The warehouses have monthly
capacities of W1 = 100 units, W2 = 40 units, and W3 = 60 units.
S1 = 30 units, S2 = 50 units, S3 = 65 units, and S4 = 55 units are
the monthly requests at the stores.
The following table lists the transportation costs from warehouses
to stores in Indian rupees:
Stores
Warehouses S1 S2 S3 S4
W1 14 16 12 20
W2 12 14 10 08
W3 10 16 08 15
PAGE 85
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Warehouses
Company A B C Supply
W 10 8 15
X 5 2 3 20
Y 6 7 4 30
Z 7 6 35
Requirement 25 26 100
Solution:
(a) x11 = 15, x21 = 10, x22 = 10, x32 = 16, x33 = 14, x43 = 35
Minimum Cost is: 753
(b) x13 = 15, x22 = 20, x33 = 30, x41 = 25, x42 = 6, x43 = 4
Minimum Cost is: 542
(c) x13 = 15, x22 = 20, x33 = 30, x41 = 25, x42 = 6, x43 = 4
Minimum Cost is: 542
Destination
Source 1 2 3 Capacity
A 42 8 10
B 11 11 30
C 10 12 28
Demand 35 40 25 100
Solution:
86 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
3.8 References
Bazaraa, M., & J. Jarvis. Linear Programming and Network Flows.
1HZ <RUN :LOH\
Bradley, S., A. Hax, & T. Magnanti. Applied Mathematical Programming.
5HDGLQJ 0DVV$GGLVRQ:HVOH\
Dantzig, G. Linear Programming and Extensions. Princeton, N.J.:
3ULQFHWRQ 8QLYHUVLW\ 3UHVV
Gass, S. Linear Programming: Methods and Applications, 5th ed.
1HZ <RUN 0F*UDZ +LOO
0XUW\ . /LQHDU 3URJUDPPLQJ 1HZ <RUN :LOH\
Wu, N., & R. Coppins. Linear Programming and Extensions. New
<RUN 0F*UDZ +LOO
Aarvik, O., & P. Randolph. “The Application of Linear Programming
to the Determination of Transmission Line Fees in an Electrical
3RZHU 1HWZRUN´ ,QWHUIDFHV
Denardo, E., U. Roth Blum, & A. Swersey. “Transportation Problem
in Which Costs Depend on Order of Arrival,” Management Science
Evans, J. “The Factored Transportation Problem,” Management Science
Gillett, B. Introduction to Operations Research: A Computer Oriented
$OJRULWKPLF$SSURDFK 1HZ <RUN 0F*UDZ +LOO
Glassey, R., & V. Gupta. “A Linear Programming Analysis of Paper
5HF\FOLQJ´ 0DQDJHPHQW 6FLHQFH
Glover, F., et al. “A Computational Study on Starting Procedures,
Basis Change Criteria and Solution Algorithms for Transportation
3UREOHPV´ 0DQDJHPHQW 6FLHQFH
88 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
4
Assignment Problem
Dr. Jigmet Wangdus
Assistant Professor
SOL/Campus of Open Learning
University of Delhi
Email-Id: jigmet@sol-du.ac.in
STRUCTURE
4.1 Learning Objectives
4.2 Introduction
4.3 Assignment Problems
4.4 Hungarian Method
4.5 Some Special Cases
4.6 Solved Illustrations (For Practical Subject)
4.7 Summary
4.8 Answers to In-Text Questions
4.9 Self-Assessment Questions
4.10 Reference
4.11 Suggested Reading
PAGE 89
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
4.2 Introduction
The assignment problem includes providing various resources (items)
to different activities (receivers) on a one-to-one basis, i.e., assign-
ing the same number of operations to an equal number of operators,
with each operator carrying out just one operation. For example, HR
department has five tasks and five Employees. The employees take
different time to complete each work depending on their efficiency.
An assignment problem is one in which one task must be assigned
to one person while minimizing the overall number of person hours.
While being a particular example of the transportation problem, the
assignment problem is not resolved by the techniques covered in
previous lessons.
To solve an assignment problem, we employ a different technique
known as the Hungarian method. Compared to other methods of deter-
mining the best solution to a transportation problem, it is simpler and
more efficient. This lesson covers a variety of assignment issue types,
including multiple solutions, maximization case, unbalance case and
restriction of assignment.
90 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Job Notes
Person 1 2…. j….. n
1 C11 C11 C1j C1n
2 C21 C22 C2j C2n
.
.
i Ci1 Ci2 Cij Cin
.
.
n Cn1 Cn2 Cnj Cnn
where Cij represents the amount of time taken by ith individual to complete
the jth job. Let xij denote the jth job assigned to the ith individual. Then
the assignment problem can be stated as follows:
m n
Minimize Z = ¦¦c x
i 1 j 1
ij ij ZKHUH L « Q DQG M « P
PAGE 91
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
4.4 Hungarian Method
One of the easiest and simple method of solving assignment problem is
Hungarian method. Hungarian method is a computer optimization approach
which solves the assignment issue in polynomial time and foreshadows
subsequent primal-dual options. Harold Kuhn coined the phrase “Hun-
JDULDQ PHWKRG´ LQ WR KRQRXU 'pQHV .QLJ DQG -HQ (JHUYiU\ WZR
Hungarian mathematicians.
Hungarian Method steps for obtaining an optimal solution of an assign-
ment problem are:
1. Determine whether the matrices are square. If it isn’t square, add
a sufficient number of dummy rows (or columns) with zero cost
or time elements to make it so.
2. Row reduction. Find the lowest cost element in every row of the
cost matrix. Take the smallest element in each row and deduct it
from each element in that row. Make sure each row has at least
one zero.
3. Column reduction. In the resulting cost matrix, find the smallest
element in each column, and then subtract it from each element in
that column. Again, make sure that each column contains at least
one zero.
4. Assign zeros.
(a) Examine the rows one at a time until you locate a row that
contains exactly one unmarked zero. Give this lone, unmarked
zero a task and encircle or [ ] it. The other zeros in the
circular zero’s column should all be crossed (×) or 0 out since
they won’t be used in any future assignments. Continue in
this way until all of the rows have been examined.
(b) Examine the rows one at a time until you locate a row that
contains exactly one unmarked zero. Give this lone, unmarked
zero a task and encircle it. The other zeros in the circular
zero’s column should all be crossed out since they won’t be
used in any future assignments. Continue in this way until
all of the rows have been reviewed.
92 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Workers
I II III IV
A 10 12 11
B 5 10 7 8
Jobs C 12 14 13 11
D 8 15 11
PAGE 93
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Solution:
Step 1: Here the matrix is square as number of rows and columns are equal.
Therefore, the given assignment problem is balanced.
Step 2: Row reduction (Find the lowest cost element in every row and
subtract this from all the elements in its row.)
Workers
I II III IV
A 0 2 1
B 0 5 2 3
Jobs C 1 3 2 0
D 0 7 3 1
Check for at least one zero in each row and each column. Otherwise go
to next step.
Step 3: Column reduction (Find the smallest element in each column and
subtract this from all the elements in its column.)
Workers
I II III IV
A 0 0 7 1
B 0 3 0 3
Jobs C 1 1 0 0
D 0 5 0 1
Since every row and column contains at least one zero, assignments can
be made.
Step 4: (Assignment)
Find the rows with exactly one zero. The first three rows contain more
than one zero. Go to Job D, where there is exactly one zero. Mark that
zero by encircle or [ ] i.e. job D is assigned to worker 1. Mark other
zeros in its column by cross (×) or Strikeout (0).
I II III IV
A 0 0 7 1
B 0 3 0 3
C 1 1 0 0
D [0] 5 1 2
94 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Step 5: Now check the columns with exactly on zero. Since there is al- Notes
ready and assignment in column I so move to column II. There is exactly
one zero. Mark the zero by encircle /[ ] and other zeros in the row by
crossed (×) or Strikeout (0).
I II III IV
A 0 [0] 7 1
B 0 3 0 3
C 1 1 0 0
D [0] 5 1 2
In column III there are more than one zero. Therefore we proceed to
column IV where there is exactly one zero. Mark the zero by encircle
/[ ] and other zeros in the row by crossed (×) Strikeout (0).
I II III IV
A 0 [0] 7 1
B 0 3 0 3
C 1 1 0 [0]
D [0] 5 1 2
Step 6: Again, examine the rows. Row B contains exactly one zero, mark
that zero by /[ ].
I II III IV
A 0 [0] 7 1
B 0 3 [0] 3
C 1 1 0 [0]
D [0] 5 1 2
Thus all the four assignments have been completed. The optimal assign-
ment schedule and total cost is:
Job Worker Cost
A II 12
B III 7
C IV 11
D I 8
Total Cost 38
The optimal assignment (minimum) cost = Rs. 38
PAGE 95
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Illustration 4.2: Solve the assignment problem. The problem has five
jobs and five machines. The costs of assignments are given below. Find
the optimal assignment schedule.
Machines
1 2 3 4 5
A 8 4 2 6 1
B 0 5 5 4
Jobs C 3 8 2 6
D 4 3 1 0 3
E 5 8 5
Solution:
Step 1: Here the matrix is square as number of rows (5) and columns
(5) are equal.
Therefore, the given assignment problem is balanced.
Step 2. Find the lowest cost element in every row and subtract this from
all the elements in its row.
Machines
1 2 3 4 5
A 7 3 1 5 0
B 0 5 5 4
Jobs C 1 6 7 0 4
D 4 3 1 0 3
E 4 0 3 4 0
Column 3 doesn’t contain zero. Move to step 3.
Step 3: Find the smallest element in each column and subtract this from
all the elements in its column.
Machines
1 2 3 4 5
A 7 3 0 5 0
B 0 4 5 4
Jobs C 1 6 6 0 4
D 4 3 0 0 3
E 4 0 2 4 0
Since every row and column contains at least one zero, assignments can
be made.
96 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
1 2 3 4 5
A 7 3 0 5 0
B [0] 4 5 4
C 1 6 6 0 4
D 4 3 0 0 3
E 4 0 2 4 0
Next, row C contains exactly one zero. So, mark the zero by encircle
and other zeros in the row by cross (×) or strikeout 0
1 2 3 4 5
A 7 3 0 5 0
B [0] 4 5 4
C 1 6 6 [0] 4
D 4 3 0 0 3
E 4 0 2 4 0
Next, Row D contains one zero. So, mark the zero by encircle /[ ] and
other zeros in the row by cross (×) or strikeout 0.
1 2 3 4 5
A 7 3 0 5 0
B [0] 4 5 4
C 1 6 6 [0] 4
D 4 3 [0] 0 3
E 4 0 2 4 0
The last row E contains more than one zero. So move to column wise.
In column 1 there is already an assignment so move to column 2. There
is exactly one zero. So, mark the zero by encircle /[ ] and other zeros
in the row by cross (×) or strikeout 0.
PAGE 97
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes 1 2 3 4 5
A 7 3 0 5 0
B [0] 4 5 4
C 1 6 6 [0] 4
D 4 3 [0] 0 3
E 4 [0] 2 4 0
In columns 3 and 4 there is an assignment so move to column 5. There
is exactly one zero. So, mark the zero by encircle /[ ] and other zeros
in the row by cross (×) or strikeout 0
1 2 3 4 5
A 7 3 0 5 [0]
B [0] 4 5 4
C 1 6 6 [0] 4
D 4 3 [0] 0 3
E 4 [0] 2 4 0
Thus all the five machines are assigned to different jobs.
The optimal assignment schedule and the total cost are:
Notes
4.5.1 Maximization Problem
A maximization problem in the form of an assignment problem could
also exist. In order to maximise the overall performance of the assign-
ment, there are situations where specific facilities must be assigned to
a number of jobs. For instance, instead of cost or times, the cells might
include profits (or any other type of revenue) that needs to be maximised.
In order to solve the problem, we must determine the opportunity loss
matrix by deducting each cell’s value from the largest value determined
from all the available cells. When a cell’s value is deducted from the
greatest value, the losses as a result of passing up an opportunity that
would have produced the highest value is revealed. The opportunity loss
matrix is the resultant matrix, and it is treated in the same way as the
minimization problem.
The maximization problem can also be solved by converting the problem
to minimum problem by multiplying all the elements Cij of the assignment
PDWUL[ E\ QHJDWLYH RQH í
Illustration 4.3: The company has to assign five agents (1, 2, 3, 4, 5)
to five different states (A, B, C, D, E). The expected revenue (in Lakhs)
for each agent is given as follows:
A B C D E
1 5 11 10 12 4
2 2 4 6 3 5
3 3 12 5 14 6
4 6 14 4 11 7
5 7 8 12 8
Since the given problem is maximization,
Step 1: Convert the problem to a minimum problem by multiply all the
elements Cij of the assignment matrix by –1
A B C D E
1 –5 –11 –10 –12 –4
2 –2 –4 –6 –3 –5
3 –3 –12 –5 –14 –6
4 –6 –14 –4 –11 –7
5 –7 ± –8 –12 –8
PAGE 99
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Step 2: Find the smallest element and subtract with others elements of
the matrix minx = – (–max Z). In the matrix the smallest value is –14.
Using this value (–14), the matrix is subtracted resulting is the min-
imization matrix. Now we can use the normal procedure to solve the
assignment problem.
A B C D E
1 3 4 2 10
2 12 10 8 11
3 11 2 0 8
4 8 0 10 3 7
5 7 5 6 2 6
Note: C1a ±± ± DQG FRQWLQXHG IRU DOO RWKHU HOHPHQWV
Step 3: Using the above matrix i.e., min Z matrix table and all cost ele-
ments non-ve. The Hungarian method can be applied to find the optimal
assignment problem.
Step 4: Row minimization
A B C D E
1 7 1 2 0 8
2 4 2 0 3 1
3 11 2 0 8
4 8 0 10 3 7
5 5 3 4 0 4
Step 5: Column minimization
A B C D E
1 3 1 2 [0] 7
2 [0] 2 0 3 0
3 7 2 0 7
4 4 [0] 10 3 6
5 1 3 4 0 3
As the number of assigned zero is less than the number of rows, we
apply next step of Hungarian method and draw the minimum number
100 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
of horizontal/vertical lines that covers all the zeroes as shown in the Notes
following matrix.
A B C D E
1 3 1 2 0 7
2 0 2 0 3 0
3 7 2 0 7
4 4 0 10 3 6
1 4 3
5 3 0
9 9 9
Now select the minimum element form amongst the uncovered elements,
which is 1 and subtract it from each of the uncovered elements and add
it to the elements which lies at the intersection of horizontal and vertical
lines. Reset of covered elements remains constant.
A B C D E
1 2 1 1 [0] 6
2 0 3 0 4 [0]
3 6 2 8 0 6
4 3 [0] 3 5
5 [0] 3 3 0 2
Again, repeat the previous steps as the number of assigned zero is less
than the number of rows.
A B C D E
1 2 1 1 [0] 6
2 0 3 0 4 [0]
3 6 2 8 0 6
4 3 [0] 3 5
3 2
5 [0] 3 0
9 9
Again, select the minimum element form amongst the uncovered elements,
which is 1 and subtract it from each of the uncovered elements and add
it to the elements which lies at the intersection of horizontal and vertical
lines. Reset of covered elements remains constant.
PAGE 101
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes A B C D E
1 2 1 [0] 0 5
2 1 4 0 5 [0]
3 6 2 7 [0] 5
4 3 [0] 8 3 4
5 [0] 3 2 0 1
A B C D E
1 32 38 40 28 40
2 40 24 28 21 36
3 41 27 33 30 37
4 22 38 41 36 36
5 33 40 35
Solution:
Since the given problem is maximization
Step 1: Convert the assignment problem into minimization problem. For
this we obtain the opportunity loss matrix by deducting each element in
the given matrix from the largest element in it. Here the largest element
is 41.
102 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
A B C D E
1 3 1 13 1
2 1 17 13 20 5
3 0 14 8 11 4
4 3 0 5 5
5 12 8 1 6 2
Step 2: Apply Hungarian method (steps) and finally obtain the following
result matrix.
A B C D E
1 8 [0] 0 7 0
2 [0] 14 12 14 4
3 0 12 8 6 4
4 1 [0] 0 5
5 11 5 0 [0] 1
A B C D E
1 8 [0] 0 7 0
2 [0] 14 12 14 4 9
3 0 12 8 6 4 9
4 1 [0] 0 5
5 11 5 0 [0] 1
Now select the minimum element form amongst the uncovered elements,
which is 4 and subtract it from each of the uncovered elements and add
it to the elements which lies at the intersection of horizontal and vertical
lines. Reset of covered elements remains constant.
PAGE 103
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Now again apply Hungarian method to the resulting matrix we get:
A B C D E
1 14 [0] 0 7 0
2 [0] 10 8 10 0
3 0 8 4 2 [0]
4 23 1 [0] 0 5
5 15 5 0 [0] 1
Thus all the five jobs are assigned to different machines.
The optimal assignment schedule and the total cost is:
Job Machines Profit
1 B 38
2 A 40
3 E 37
4 C 41
5 D 35
Total Profit
Solution: Notes
Step 1: Since the matrix is not a balanced square matrix, convert the
above 4 × 5 matrix into a square matrix by adding one dummy row D5.
Machines
A B C D E
1 5 7 11 6 7
Job
2 8 6 5 6 5
3 6 7 10 7 3
4 10 4 8 2 4
D5 0 0 0 0 0
A B C D E
1 0 2 6 1 2
2 3 1 0 1 0
3 3 4 7 4 0
4 8 2 6 0 2
D5 0 0 0 0 0
A B C D E
1 [0] 2 6 1 2
2 3 1 [0] 1 0
3 3 4 7 4 [0]
4 8 2 6 [0] 2
D5 0 [0] 0 0 0
PAGE 105
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Thus all the five jobs are assigned to different machines.
The optimal assignment schedule and the total cost is:
Machines
L M N O
A 18 24 28 32
Job
B 8 13 17
C 10 15 22
Solution:
Since the above matrix is not a balanced square matrix,
Step 1: Convert it into square matrix by introducing one dummy row D.
Machines
L M N O
A 18 24 28 32
Job
B 8 13 17
C 10 15 22
D 0 0 0 0
106 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Step 2: Row reduction (Find the lowest cost element in every row and Notes
subtract this from all the elements in its row.)
L M N O
A 0 6 10 14
B 0 5 11
C 0 5 12
D 0 0 0 0
L M N O
A [0] 6 10 14
B 0 5 11
C 0 5 12
D 0 [0] 0 0
L M N O
A [0] 6 10 14
B 0 5 11
C 0 5 12
D 0 0 0 0
Now select the minimum element form amongst the uncovered elements,
which is 5 and subtract it from each of the uncovered elements and add
it to the elements which lies at the intersection of horizontal and vertical
lines. Reset of covered elements remains constant.
PAGE 107
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Now again apply Hungarian method to the resulting matrix we get:
L M N O
A [0] 1 5
B 0 [0] 4 6
C 0 0 4 7
D 5 0 [0] 0
Again, as the number of assigned zero is less than the number of rows,
repeat the above steps.
L M N O
A [0] 1 5
B 0 [0] 4 6
C 0 0 4 7
D 0 0 [0] 0
Modified matrix.
L M N O
A 0 1 1 5
B 0 0 0 2
C 0 0 0 3
D 4 0 0
L M N O
A [0] 1 1 5
B 0 [0] 0 2
C 0 0 [0] 3
D 4 0 [0]
108 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
The optimal assignment schedule and the total cost is: Notes
Solution 2:
L M N O
A [0] 1 1 5
B 0 0 [0] 2
C 0 [0] 0 3
D 4 0 [0]
Notes Illustration 4.7: The cost of performing different jobs by different per-
sons is given below. Some individuals cannot perform specific jobs, they
are marked X. Using this information state (i) the optimal assignment of
jobs, and (ii) the cost of such assignment.
Job
1 2 3 4 5
Person A 27 18 X 20 21
B 31 24 21 12 17
C 20 17 20 X 16
D 22 28 20 16 27
Solution
Step 1: Since the above matrix is not a squared (balanced) matrix, covert
it by introducing a dummy row (Z).
We assign M to the cells for which no preference is given.
Job
1 2 3 4 5
A 27 18 M 20 21
B 31 24 21 12 17
Person
C 20 17 20 M 16
D 22 28 20 16 27
Z (Dummy) 0 0 0 0 0
110 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
1 2 3 4 5
A 0 M 2 3
B 12 0 5
C 4 1 4 M 0
D 6 12 4 0 11
X (Dummy) 0 0 0 0 0
1 2 3 4 5
A [0] M 2 3
B 12 [0] 5
C 4 1 4 M [0]
D 6 12 4 0 11
X (Dummy) 0 0 0 0 0
Now select the minimum element form amongst the uncovered elements,
which is 4 and subtract it from each of the uncovered elements and add
it to the elements which lies at the intersection of horizontal and vertical
lines. Reset of covered elements remains constant.
PAGE 111
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Now again apply Hungarian method to the resulting matrix we get:
1 2 3 4 5
A [0] M 6 3
B 15 8 5 [0] 1
C 4 1 4 M [0]
D 2 8 [0] 0 7
X (Dummy) [0] 0 0 4 0
Now each row and column has one and only one assigned zero, the as-
signment with maximum satisfaction is made.
The optimal assignment schedule and the total cost is:
112 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Solution: Notes
Step 1: Since the above matrix is not a square matrix, we introduce a
dummy column d with cost and time equal to zero.
Men
1 2 3 d
W 26 15 0
Task X 13 27 6 0
Y 35 20 15 0
Z 18 30 20 0
Step 2: Row reduction: Since each contains minimum one zero move
to next step.
Step 3: Column reduction:
1 2 3 d
W 0 6 0
X 4 7 0 0
Y 26 0 0
Z 10 14 0
Step 4: Assignment
1 2 3 d
W [0] 6 0
X 4 7 [0] 0
Y 26 [0] 0
Z 10 14 [0]
The optimal assignment schedule and the total cost is:
Task Men Cost
W 1
X 2 6
Y 3 20
Z D 0
Total Cost 35
Illustration 4.9: To generate interest and motivate students towards
the area of finance, the college decides to hold special webinar on four
topics - leasing, portfolio management, private mutual funds, swaps, and
PAGE 113
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes options. Such webinar would be held only once per week. However, the
scheduling of these webinars has to be done carefully so that the number
of students unable to attend is kept to a minimum. A careful study indi-
cates that the number of students who cannot attend a particular seminar
on a specific day is given below:
Webinar
Portfolio Private Mutual
Leasing Management Funds Swaps
Monday 50 40 60 20
Tuesday 40 30 40 30
Day Wednesday 60 20 30 20
Thursday 30 30 20 30
Friday 10 20 10 30
Find the optimal schedule of the webinar. Also find the total number of
students who will be missing at least one seminar.
Solution:
Step 1: Since the above matrix is not a balanced square matrix, convert
it into balanced matrix by introducing a dummy column D.
Webinar
Portfolio Private
Leasing Management Mutual Funds Swaps D
Monday 50 40 60 20 0
Tuesday 40 30 40 30 0
Day
Wednesday 60 20 30 20 0
Thursday 30 30 20 30 0
Friday 10 20 10 30 0
114 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Step 4: Assignment
Portfolio Private
Leasing Management Mutual Funds Swaps D
Monday 40 20 50 [0] 0
Tuesday 30 10 30 10 [0]
Wednesday 50 [0] 20 0 0
Thursday 20 10 10 10 0
Friday [0] 0 0 10 0
Portfolio Private
Leasing Management Mutual Funds Swaps D
Monday 40 20 50 [0] 0
Tuesday 30 10 30 10 [0]
Wednesday 50 [0] 20 0 0
Thursday 20 10 10 10 0
Friday [0] 0 0 10 0
PAGE 115
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Now select the minimum element form amongst the uncovered elements,
which is 10 and subtract it from each of the uncovered elements and add
it to the elements which lies at the intersection of horizontal and vertical
lines. Reset of covered elements remains constant.
Now again apply Hungarian method to the resulting matrix we get:
Portfolio Private
Leasing Management Mutual Funds Swaps D
Monday 30 20 40 [0] 0
Tuesday 20 10 20 10 [0]
Wednesday 40 [0] 10 0 0
Thursday 10 10 [0] 10 0
Friday [0] 10 0 20 10
Now each row and column has one and only one assigned zero, the as-
signment with maximum satisfaction is made.
The optimal assignment schedule and the number of students absent is:
IN-TEXT QUESTIONS
1. The cost elements in the matrix of a typical assignment problem
indicate the cost of performing the particular jobs by particular
personnel. (True/False)
2. An assignment problem is said to be balanced when the number
of rows in the given matrix matches with the number of column.
(True/False)
3. An assignment problem is always solved on the assumption that
the assignments are to be done on a one-to-one basis. (True/False)
116 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
4.7 Summary
Assignment problem deals with assignment of persons to jobs that
they can perform with varying efficiency.
The solution to an assignment problem can be obtained by complete
enumeration and evaluation of all possible assignments. However the
method becomes impractical for even moderately large values of n.
An unbalanced problem has to be balanced first by introducing the
required number of dummy workers/jobs.
For every prohibited assignment, the given cost element is replaced
by M, which is a very large value.
Multiple zeros in all columns and rows are indicative of multiple
optimal solutions.
1. True 5. True
2. True 6. True
3. True 7. True
4. True 8. True
PAGE 117
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
4.9 Self-Assessment Questions
1. What is an assignment problem?
2. What is an unbalanced assignment problem? How is the Hungarian
method applied in respect of such problem?
3. How would you deal with the assumption problems where
(a) Some assignments are prohibited.
(b) The objective function is of maximization type
4. A firm employs five typists on hourly piece-rate basis for their
daily work on five different jobs. The charges of each typist per
job are given below:
Job
A B C D E
1 85 75 65 125 75
2 78 66 132 78
Typist 3 75 66 57 114
4 80 72 60 120 72
5 76 64 56 112 68
Find the least cost allocation. Sol. (75 + 66 + 66 + 80 + 112 = 399)
5. ABC company has four salespersons who are assigned four different sales
territories. The monthly sales increase estimated for each salesperson
for different sales territories are shown in the table given below:
Sales Territories
Salesperson I II III IV
A 200 150 170 220
B 160 120 150 140
C 200
D 180 175 160
(a) Suggest optimal assignment and total maximum sales increase
per month.
(b) If for certain reasons, salesperson B cannot be assigned to sales
territory III will the optimal assignment schedule be different?
If so find that schedule and the effect on total sales.
Sol. (a) (220 + 160 + 190 + 175 = 745)
(b) (220 + 150 + 195 + 180 = 745)
118 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
6. A firm produces four products. There are four operators who are Notes
capable of producing any of these four products. The profits for
each of the products are given below:
Products
Operators W X Y Z
1 240 300 225 300
2 360 450 250 300
3 144 180 150 200
4 240 300 225 180
Find the optimal assignment of products to operators.
Sol. (Profit 300 + 450 + 150 + 240 = 1,140)
4.10 Reference
Vohra, N. D. 2012. Quantitative Techniques in Management,
4th edition, Tata McGraw Hill Education Private Limited.
PAGE 119
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
5
Project Scheduling
Dr. Sandeep Mishra
Assistant Professor
Shaheed Rajguru College of
Applied Sciences for Women
University of Delhi
Email-Id: sandeepstat24@gmail.com
STRUCTURE
5.1 Learning Objectives
5.2 Introduction: Project Scheduling
5.3 Construction of AOA Network Diagram
5.4 Scheduling with Known Activity Times
5.5 Scheduling with Uncertain Activity Times
5.6 7LPH&RVW 7UDGH2ৼV
5.7 Summary
5.8 Answers to In-Text Questions
5.9 Self-Assessment Questions
5.10 References
5.11 Suggested Reading
120 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
5.2 Introduction: Project Scheduling
Have you ever overseen a significant event? You might have served as
the prom committee chair or the board chair for the graduation ceremony
in high school. You might have led your team during the introduction of
a new product, the planning of a facility expansion, or the implementa-
tion of enterprise resource planning. Even if you have never managed
people in such circumstances, you have undoubtedly had your own per-
sonal projects to contend with, such as writing a paper, moving to a new
apartment, applying to college, or selling a house. As a volunteer, you
may have overseen the annual function, the elementary school picnic, or
the river clean-up project. How did you plan your day’s events? Most of
your projects, did they finish on time? How did you handle unforeseen
circumstances? Did you finish your work on time and on budget? All of
these are crucial components of project management. Project managers
with expertise are essential assets for organisations since they handle
projects frequently.
A project involves many interrelated activities (or tasks) that must be
completed on or before specified time limit, in a specified sequence (or
order) with specified quality and minimum cost of using resources such
as personnel, money, material, facilities and/or space.
In this lesson we mainly focus on creating and managing schedule. This
covers project scheduling with known activity times using well known
techniques - PERT and CPM, scheduling with uncertain activity and trade
cost trade-offs.
PAGE 121
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes head event (j) represents the completion of the activity as shown in Fig..
The activities can be further classified into the following three categories:
(i) Predecessor Activity: A predecessor activity is one that must be
finished before one or more other actions may begin.
(ii) Successor Activity: A successor activity is one that begins right
away after the completion of one or more other activities.
(iii) Dummy Activity: Dummy activity is a term used to describe an
activity that doesn’t use any resources or time.
Only the given precedence relationship between the project’s activities is
added as a fake activity to the network. It is required in situations when
(a) two or more concurrent project activities have the same head and tail
events, or (b) two or more activities share some—but not all—of their
immediate preceding activities. In the network diagram depicted in the
Figure, a dotted line represents a fictitious activity.
To illustrate the order of importance of the project’s activities, network
models use the two precedence networks listed below.
Activity-on-Node (AON) Network: In this kind of precedence network,
each node (or circle) stands for a particular task, and the arcs show the
hierarchy of tasks. The activities are located within the nodes in AON
network diagrams, and the arrows are used to denote the necessary
sequencing. These diagrams typically don’t feature beginning and ending
nodes for the entire project. These diagrams are always simpler to draw
and understand because they don’t include any dummy activity.
Activity-on-Arrow (AOA) Network: In this type of precedence network
at each end of the activity arrow is a node (or circle). These nodes stand
for instants or points in time when an activity begins or ends. The length
of time needed to complete that action is indicated by the arrow itself. In
these diagrams, any actions that have no predecessors may begin at the
same beginning node. The diagram then moves from left to right before
coming to a finish at a single node where all of the unfollowed activities
are gathered. The following are three key benefits of employing AOA:
(i) Many computer programs are based on AOA network.
(ii) AOA diagrams can be superimposed on a time scale with the arrows
drawn, the correct length to indicate the time requirement.
(iii) AOA diagrams give a better sense of the flow of time throughout a
project.
122 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 123
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
124 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
A wide range of projects can be planned, scheduled, and managed using Notes
PERT and CPM:
Research and development of new products and processes
Construction of plants, buildings, and highways
Maintenance of large and complex equipment
Design and installation of new systems
Project managers are responsible for planning and coordinating the numer-
ous tasks or activities in these kinds of projects to ensure that everything
is finished on time.
PAGE 125
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
5.4.2 Critical Path Analysis
The objective of critical path analysis is to predict the project’s overall
duration and give starting and finishing durations to every activity
involved. This makes it easier to compare the project’s actual progress
to its projected completion date.
The expected duration of an activity is estimated from the duration of
individual activities, which may be determined uniquely (in the case of
CPM) or may entail three-time estimates (in the case of PERT). The fol-
lowing elements need to be understood in order to establish the project
scheduling.
(i) Total completion time of the project.
(ii) Earlier and latest start time of each activity.
(iii) Critical activities and critical path.
(iv) Float for each activity.
Notations:
Ei Earliest occurrence time of an event, i. This is the latest time for
an event to occur when all the preceding activities have been completed,
without delaying the entire project.
Li Latest allowable time of an event, i. This is the latest time at which
an event can occur without causing a delay in project’s completion time.
ESij Early starting time of an activity (i, j).
LSij Late starting time of an activity (i, j).
EFij Early finishing time of an activity (i, j).
LFij Late finishing time of an activity (i, j).
tij duration of an activity (i, j).
There should only be one start event and one finish event in a project
schedule. The other events are numbered consecutively with integer 1,
2, …, n, such that i j for any two events i and j connected by an
activity, which starts at i and finishes at j.
This schedule for each activity is created using a two-pass approach
that includes a forward pass and a backward pass. The earliest times
126 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
( ESij and EFij ) are determined during the forward pass. The latest times Notes
( LSij and LFij ) are determined during the backward pass.
PAGE 127
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes time for an event at the beginning of the activity minus the activity
duration. Mathematically,
Total Float TFij = L j Ei tij = LSij í ESij LFij í EFij
(b) Free Float: This is the amount of time that each non-critical ac-
tivity’s completion time can be pushed back without impacting its
immediately succeeding activities. The amount of free float time for
a non-critical activity (i, j) is computed as follows:
Free Float FFij = E j Ei tij
Min ^ESij , for all immediate successor of activity i, j ` EFij
(c) Independent Float: This is the length of time that any non-critical
activity i j can be delayed without affecting the completion times
of the activities that come before or after it. Each non-critical activ-
ity’s independent float time is calculated mathematically as follows:
Independent Float IFij = E j Li tij = ^ES ij LSij ` tij
Independent float values that are negative are regarded as zero.
128 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Duration Immediate
Activity Description
(weeks) Predecessors
A Design new premises 14 -
B Obtain tenders from the contractors 4 A
C Select the contractors 2 B
Arrange details with selected
D 1 C
contractors
E Decide which equipment is to be used 2 A
F Arrange storage of equipment 3 E
G Arrange disposal of other equipment 2 E
H Order new equipment 4 E
I Take delivery of new equipment 3 H, L
J Renovations take place 12 K
Remove old equipment for storage
K 4 D, F, G
or disposal
Cleaning after the contractor has
L 2 J
finished
M Return old equipment for storage 2 H, L
(a) Draw the network diagram showing the interrelations between the
various activities of the project.
(b) Calculate the minimum time that the renovation can take from the
design stage.
PAGE 129
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes (c) Find the effect on the overall duration of the project if the estimates
or tenders can be obtained in two weeks from the contractors by
reducing their numbers.
(d) Calculated the independent float that is associated with the non-critical
activities in the network diagram.
Solution: (a) The network diagram for the given project, along with
E-values and L-values, is shown in Fig 1.
The critical path in the network diagram (Figure 5.1) has been shown by
bold lines joining all those events where E-values and L-values are equal.
(b) The critical path of the project is: 1 2 3 4 7 8 9 10 12 and
critical activities are A, B, C, D, K, J, L and I. The total project com-
pletion time is 42 weeks. For non-critical activities, the total float, free
float and independent float calculations are shown in Table 5.1.
(c) The effect on the overall project duration, if the time of activity B is
reduced to 2 weeks instead of 4 weeks, is shown in Table 5.2.
130 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
IN-TEXT QUESTIONS
1. The objective of the project scheduling is to minimize total
project cost. (True/False)
2. The CPM is used for completing the projects that involves
activities of repetitive nature. (True/False)
3. PERT is referred to as an activity-oriented technique. (True/
False)
4. _____________is the time-consuming job or task that is a key
subpart of the total project.
§ TS Te ·
prob ¨ Z ¸
© Vi ¹
132 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 133
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
134 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Probability that the project will be completed at least 4 weeks earlier (i.e.
13 weeks) than the expected project duration of 17 weeks is given by
T T 13 17 ½
Prob. ® Z d s e ¾ Prob.^Z d 1.33` 0.5 0.4082 0.09188
¯ V 3 ¿
Thus the probability of completing the project in less than 13 days is
IN-TEXT QUESTIONS
5. Beta probability distribution is often used in computing the
expected activity completion times and variances in networks.
(True/False)
6. The shortest possible time (duration) in which an activity can be
performed assuming that everything goes well is _______________.
7. The amount of time that is expected to complete the activity
is called ____________________.
PAGE 135
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes the length of a project in the most affordable way possible. Additionally,
extending an activity’s duration past its usual point (cost-efficient) may
raise the expense of carrying out that action. For the sake of simplicity,
it is assumed that the relationship between an activity’s normal time and
cost as well as crash time and cost is linear. Therefore, by calculating
the relative change in the cost per unit change in time, the crash cost
per unit of time may be determined.
The values of cost slope for critical activities indicate the direct extra
cost required to execute an activity per unit of time.
Step 3: For reducing the total project completion time, identify and crash
an activity time on the critical path with lowest cost slope value to the
point where
(i) another path in the network becomes critical, or
(ii) the activity has been crashed to its lowest possible time.
Step 4: If the critical path under crashing is still critical, return to step
3. However, if due to crashing of an activity time in step 3, other path(s)
in the network also become critical, then identify and crash the activity(s)
on the critical path(s) with the minimum joint cost slope.
Step 5: Terminate the procedure when each critical activity has been
crashed to its lowest possible time. Determine total project cost corre-
sponding to different project durations.
136 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Example 3: The data on normal time, cost and crash time and cost Notes
associated with a project are shown in the following table.
PAGE 137
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes (b) The normal total project cost associated with normal project duration
of 32 weeks is as follows:
Total cost = Direct normal cost + Indirect cost for 32 weeks
= 4,220 + 50 × 32 = Rs. 5,820
(c) Calculations for total float associated with non-critical activities are
shown in Table 5.4.
Table 5.4: Total Float
The minimum value of crash cost per week is for activity 2 – 5 and
±+HQFHFUDVKLQJDFWLYLW\±E\GD\VIURPZHHNVWRZHHNV
But the time should only be reduced by 1 week otherwise another path
1 2 3 5 6 7 8 becomes a parallel path. Network, as shown in Figure 5.4,
is developed when it is observed that new project time is 31 weeks and
the critical path are 1 2 5 6 7 8 and 1 2 3 5 6 7 8 .
138 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
With crashing of activity 2 – 5, the crashed total project cost becomes: Notes
Crashed total cost = Total direct normal cost + Increased direct cost
due to crashing of activity (2 – 5) + Indirect cost for 31 weeks
= 4,220 + 1 × 45 + 50 × 31 = 4,265 + 1,550 = Rs. 5,815
For revised network shown in Figure 5.4, new possibilities for crashing
critical activities are listed in Table 5.6.
Since crashed cost slope for activity 5 – 6 is minimum, its time may
be crashed by 2 weeks from 6 weeks to 4 weeks. The updated network
diagram is shown in Figure 5.5.
PAGE 139
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
It may be noted in Figure 5.5, that the critical paths shown in Figure
5.4 remain unchanged because activity 5 – 6 is common in both. With
crashing of activity 5 – 6 by 2 weeks, the crashed total cost becomes:
Crashed total cost = Total direct normal cost + Increased direct cost
GXH WR FUDVKLQJ RI DFWLYLW\ ± ,QGLUHFW FRVW IRU ZHHNV
î î î 5V
For revised network given in Figure 5.5, new possibilities for crashing
in the critical paths are listed in Table 5.7.
Table 5.7: Crash Cost Slope
140 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
WLPH+HQFHWHUPLQDWHFUDVKLQJ7KHRSWLPDOSURMHFWGXUDWLRQLVZHHNV Notes
with associated cost of Rs. 5,805 as shown in Table 5.8.
IN-TEXT QUESTIONS
8. The process of shortening the duration of a project in the least
expensive manner possible is called ____________________.
9. In time-cost trade-off function analysis the:
(a) Cost decreases linearly as time increases
(b) Cost at normal time is zero
(c) Cost increases linearly as time decreases
(d) None of the above
5.7 Summary
The principles of project management approaches are covered in this
lesson. We talk about two methods, PERT and CPM, both of which are
very useful for managing big, complicated projects. Project networks
are used to express projects in the beginning. We can then determine
the project schedule and the critical path using a two-pass technique(s).
PERT is probabilistic and provides for three times estimations for each
activity; the expected completion time and variance of the project are
calculated using these estimates. We demonstrate how to use these pa-
rameters to calculate the likelihood that the project will be finished by
a particular date. We go over how to plan, schedule, track, and manage
project expenditures using project management approaches. Finally, we
go over how to “crash” projects, which involves using more resources
and the trade-off method to speed up their completion time.
PAGE 141
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
5.8 Answers to In-Text Questions
1. False
2. True
3. False
4. Activity
5. True
6. Optimistic time
7. Most likely time
8. Crashing
9. (a) Cost decreases linearly as time increases
5.10 References
Balakrishnan, N., Render, B., Stair, R. M., & Munson, C. (2017).
Managerial decision modelling. Upper Saddle River, Pearson
Education.
Hillier, F., & Lieberman, G. J. (2014). Introduction to operations
research (10th ed.). McGraw Hill Education.
142 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
6
Decision Theory
Dr. Abhishek Kumar Singh
Assistant Professor
PGDAV(E) College
University of Delhi
Email-Id: abhishekbhu008@gmail.com
STRUCTURE
6.1 Learning Objectives
6.2 Introduction
6.3 'H¿QLWLRQ RI 'HFLVLRQ 7KHRU\
6.4 'H¿QLWLRQ RI 'HFLVLRQ 0DNLQJ (QYLURQPHQW
6.5 Statistical Decision Theory
6.6 Elements and Steps of a Decision
6.7 The Principles and Approach of Decision Theory
6.8 3D\Rৼ 7DEOH DQG ([SHFWHG 3D\Rৼ
6.9 Opportunity Loss
6.10 Importance of Accurate Information
6.11 Sensitivity Analysis and Decision Trees
6.12 Answers to In-Text Questions
6.13 Self-Assessment Questions
6.14 References
6.15 Suggested Readings
Notes
6.2 Introduction
This chapter facilitates an overview of decision theory is to support
individuals and organisations in their decision-making. It offers a useful
conceptual framework for making crucial decisions. Making a decision
involves choosing the course of action that, in the opinion of the decision-
maker, is the best under the circumstances from among a variety of
alternatives.
In this lesson, you will be able to understand the situations, options for
making decisions, and outcomes of those actions. You will also learn
about how to arrange data in a decision tree or reward table. Determine
the expected return on the alternative decision, Assess the information’s
forecasted value and the existing and anticipated opportunity losses.
144 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
the issue can be broken down into different courses of action, natural states, Notes
potential outcomes, and anticipated rewards from each of these possibilities.
Let’s go over the ideas behind the choice theory method of issue solving.
The payoffs required for each decision alternative and natural state Notes
combination.
PAGE 147
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
and have a monetary value. In this article, we discover that there are Notes
positive risks (opportunities) and negative risks (threats). By examining
the anticipated monetary value, the decisions to be made are explored.
How much money you will have in the future depends on some of your
decisions. The determinants in this issue are the likelihood of an outcome
occurring and its likelihood of occurring. (Be the expected monetary value
for the decision alternative Ai. In this essay, we cover expected monetary
value, including its definition, the formula, and how to calculate EMV.)
EMV is utilized in the process of performing a quantitative risk
analysis and is a crucial component of risk management.
As this method requires mathematical computations, many PMP
candidates overlook it. I do not recommend avoiding it. There is only
one EMV formula needed for this idea.
EMV(Ai >P(Sj) · V(Ai, Sj)
Assign Ai as the ith possible decision.
Define P(Sj) as the likelihood of the jth natural state.
Define V(Ai, Sj) as the payoff value for the decision option Ai and
natural state Sj pair.
Assume that the decision alternative Ai’s expected monetary value
is EMV (Ai).
Example 1: You recognise a danger connected to the market claim while
managing a software project for an IT company. There is a 20% chance
of risk, and if it materialises, you will lose $8000. We shall now deter-
mine the EMV of this risk.
20% chance of the event occurring
Solution: Impact of risk: – $8000
EMV = P u I
EMV = 0.2 u –8000 = –$1600
Let’s first talk about likelihood and impact since these are factors in the
EMV calculation.
Probability
The probability of any event happening is known as probability. A coin
flip, for instance, has a 50% chance of coming up heads and a 50% chance
PAGE 149
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes of coming up tails. You, therefore, claim that there is a 50% chance of
seeing heads or tails. We shall now talk about it mathematically.
The probability calculation formula is as follows: The likelihood of an
event occurring is equal to the number of positive events that could take
place (Total number of events). Let’s see how the aforementioned formula
relates to the coin Example.
There were 2 events overall (because the coin can either show heads or
tails).
Total number of favourable circumstances = 1 (assumes heads will show).
The likelihood of a head is given by (the number of favourable events)/1.
(Total number of events) = 1/2 = 50%
Hence, if you flip a coin, there is a 50% chance that it will display heads.
Example 2: Your team has identified three risks with probabilities of
10%, 50%, and 35% throughout the risk management planning process.
The first two risks will cost you 5,000 USD and 8,000 USD, respective-
ly, whilst the third risk will reward you with 10,000 USD. Calculate the
anticipated financial impact of these risk events.
Solution: The expected financial value of the three events is equal to
the EMVs of the first, second, and third occurrences. The first event’s
EMV was 0.10 * (–5,000) = –500.
The second event’s EMV is 0.50 * (–8,000) = –4,000
Third event’s EMV equals 0.35 u 10,000 = 3,500 EMV of all three events
is the sum of the EMVs of the first, second, and third occurrences.
= – 500 – 4,000 + 3,500 = –1,000
All three events have an expected monetary value (EMV) of $1,000.
Advantages from EMV Analysis
It provides you with the average result of all the noted uncertain
events.
It aids in the contingency reserve calculation.
It aids in making the optimal decision in a decision tree analysis.
It merely needs the advice of professionals; it does not need any
expensive resources.
It supports your make-or-buy choice when it comes to plan purchasing.
150 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Where
EOL (Ai) refers to the expected opportunity loss for a particular decision
alternatives.
P(Sj) refers to the probability associated with the states of nature j.
R(Ai, Sj) refers to the regret or loss for a particular combination of a
state of nature and a decision alternative.
PAGE 151
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Calculation of Opportunity Loss – Simply deduct the actual payoff amount
from the ideal payoff amount to determine the probable opportunity loss.
Opportunity Loss
.60 Market .40 Market Expected
Purchase Rise Decline Opportunity loss
Kayser Chemicals $0 $150 $60
Rim Homes 200 50 140
Texas Electronics 500 0 300
(A1) = (.6) ($0) + (.4) ($150) = $60
(A1) = (.6) ($200) + (.4) ($50) = $140
(A1) = (.6) ($500) + (.4) ($0) = $300
152 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
6.10 Importance of Accurate Information
What Value does Information that is known before a Strategy is used
have?
The difference between the expected reward if the state of nature were
known and the best choice under the circumstances of ambiguity is known
as the Expected Value of Perfect Information (EVPI).
The price one would be prepared to pay to obtain perfect information is
known as the anticipated value of perfect information (EVPI) in deci-
sion theory. [1] Health economics is one area where the EVPI concept
is frequently used. There is always a degree of uncertainty surrounding
a decision in that situation and while considering whether to adopt a
new treatment method because there is always a chance that the decision
will turn out to be incorrect. The expected value of perfect information
analysis seeks to quantify the anticipated expense of that uncertainty,
which “can be interpreted as the expected value of perfect information
(EVPI), since perfect information can eliminate the possibility of making
the incorrect decision,” at least theoretically.
The effect on decisions when more information is provided so that the
decision-maker can arrive at a point of certainty is referred to as the
expected value of perfect information (EVPI). Decision-makers might
use EVPI as a guide to determine how much to spend on further infor-
mation. The decision-maker can benefit from this additional information
by getting this insight, whether it relates to a better grasp of the market
or the expenses of changing the company. When the decision-makers do
make a choice, they can do so with more confidence in the anticipated
outcomes.
A company that wishes to buy a piece of another company, for instance,
would probably decide to do so without a second thought if the deci-
sion-makers had complete knowledge of the possible outcomes. The firm
can use the expected value of perfect information to select the best option
from the available options in order to boost the certainty of their decision.
Calculate the EVPI Formula
Use the general formula: profit from perfect information x likelihood of
that result to calculate EVPI. We can rely on the equation EVPI = EV
PAGE 153
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
154 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 155
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes alternatives, natural state probabilities, and payoffs for any combination
of these variables. The decision-making point is indicated by a box, and
the choices being considered are indicated by the branches coming out of
the box. A decision tree is a visual representation of the decision-making
process that shows option alternatives, natural state probabilities, and
payoffs for any combination of these variables. The decision tree analy-
sis criterion that is most frequently utilised is expected monetary value
(EMV). One of the initial steps in this type of analysis is to graph the
decision tree and to define the financial repercussions of every possible
solution for a certain problem.
IN-TEXT QUESTIONS
1. Which section of the lesson discusses the definition of decision
theory and the decision-making environment?
(a) Introduction
(b) Statistical Decision Theory
(c) Payoff table and expected payoff
(d) Sensitivity analysis and decision trees
2. What is the purpose of statistical decision theory?
(a) To analyze historical data
(b) To predict future outcomes
(c) To evaluate decision-making environments
(d) To minimize decision-making risks
156 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 157
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
158 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
6.13 Self-Assessment Questions
I. Short Answer Questions:
1. Define Decision Theory and explain with examples.
2. What are the examples of decision-making questions?
3. What are the four elements of decision theory?
4. What are the key issues in decision theory?
5. What are the six steps of Decision Theory?
6. What do understand by the term Decision-making environment?
7. Define Decision Tree.
PAGE 159
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
(a) Finish the payout table. [Showing work for this point is optional]
(b) How many copies of Waldo Books should the company print
using the most optimistic scenario?
160 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
6.14 References
“Definition and significance of decision theory.” Dictionary.com.
obtained on April 2, 2022.
0U5RJHU%0\HUVRQ ³)RXQGDWLRQDO,GHDVLQ'HFLVLRQ
Theory.” Conflict is analysed using game theory. Harvard University
3UHVV ,6%1 &DPEULGJH 0DVVDFKXVHWWV
Levchenko A., Lipniacki T., Habibi I., Cheong R., Emamian E. S.,
& Abdi A. (April 2017). “Cell decision making error computation
and measurement utilising single cell data.” e1005436. PLOS
&RPSXWDWLRQDO %LRORJ\ 30,' %LEFRGH
PLSCB..13E5436H. doi:10.1371/journal.pcbi.1005436. Obtained on
April 2, 2022.
Decision theory: A basic introduction, Sven Ove Hansson. A truly
trans disciplinary subject, 2005 Section 1.2.
162 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
7
Decision Under Conflict
Dr. Shubham Agarwal
Associate Professor
New Delhi Institute of Management
GGSIP University
Email-Id: meshubhamagarwal@gmail.com
STRUCTURE
7.1 Learning Objectives
7.2 Introduction
7.3 Types of Games
7.4 Strategy
7.5 3D\RৼV 0DWUL[
7.6 Saddle Point, Optimal Strategies and Value of the Game
7.7 Minimax (Maximin) Principle
7.8 Game without Saddle Point
7.9 Solution of 2 u 2 Games without Saddle Point
7.10 Dominance Rule
7.11 Graphical Method to Solve 2 u n, m u 2 and 2 u 2 Games
7.12 Applications of Game Theory
7.13 Limitations of Game Theory
7.14 Summary
7.15 Answers to In-Text Questions
7.16 Self-Assessment Questions
7.17 References
7.18 Suggested Readings
PAGE 163
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
7.1 Learning Objectives
Explain the term game theory.
Identify strategic situations and represent them as games.
Solve simple games using various techniques.
Analyse economic situations using game theoretic techniques.
Recommend and prescribe which strategies to implement.
Explain game theory as an abstraction of everyday decision making
in the business world.
7.2 Introduction
The area of decision theory that is concerned with interdependent
decisions is game theory. The challenges of interest involve a number of
parties, each of whom has particular goals connected to a shared system
or resources. The issues are known as games, and the participants are
known as players because game theory was developed via the examination
of competitive scenarios. These methods, however, are not exclusive to
sporting events and can be used in non-competitive settings as well. Game
theory is described as “the mathematical study of strategies for dealing
with competitive situations where the outcome of a participant’s choice of
action vitally depends on the actions of other participants” in the Com-
pact Oxford English Dictionary. A game consists of three components:
A set of players
A set of available strategies for each player
A set of payoffs to each player for each possible configuration of
strategies
Consider the two players in a chess game. The collection of moves that
a player can make at each stage of the game, as specified by the chess
rules, makes up the potential tactics. Pay-outs include the outcomes “Win,
Lose,” “Lose, Win,” and “Draw, Draw,” wherein each brace includes the
game’s outcome for player number 1 and its corresponding outcome for
player 2. The sequences of tactical decisions (moves) made by the two
players over the entire game determine the outcomes once more using
the chess rules.
164 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
The Theory of Game and Economic Behaviour, which Neumann and Notes
0RUJHQVWHUQ LQWURGXFHG LQ XQGHUOLQHG WKH EDVLF WHUPLQRORJLHV DQG
issue formulation. Decisions in which the results depend on the partic-
ipants’ strategies are the focus of game theory. Game theory offers a
method for deriving the basic framework of interaction and representing
it as a tactical game.
PAGE 165
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
7.3.4 Simultaneous and Sequential Games
A game can either involve sequential moves (or decisions), as in the case
of chess, or simultaneous decisions, as in the case of the kid-friendly game
rock-paper-scissors. As we shall see, the distinction between simultaneous
and sequential games is more about the information accessible to players at
the moment of a move than it is about the timing of the moves (whether
moves are made at the same time or at different times). A player in a
sequential game knows which specific decision his opponent made out
of all the options open to him, whereas in a simultaneous game, players
make decisions before learning about the decision made by the other.
Notes
7.3.7 Infinitely Long Games
Games are often completed in a finite number of moves, according to
economists and players of real-world games. Set theorists in particular
investigate games with an endless number of moves, with the winner (or
other payoff) not known until all of those moves have been completed.
Pure mathematicians are not bound in this way.
7.4 Strategy
A strategy is one of the predetermined possible actions of a participant
in a game played strategically. A strategy in a lengthy game is an entire
set of options, one for each choice the player must make.
PAGE 167
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
[minj (vij)] = minj [maxi (vij)] = vrs, then the matrix is said to have a Notes
saddle point (r, s).
If the payoff matrix [vij] has a saddle point (r, s), then the player A and
B are said to have rth and sth optimal strategies respectively.
The payoff vrs at the saddle point (r, s) is called the value of the game and
it is obviously equal to the minimax and maximum value of the game. A
game is said to be fair if, minimax value = maximum value = 0. A game
is said to be strictly determinable if, minimax value = maximum value = v.
Player B
I II III
I 3 2 6
Player A
II 2 0 2
III 5 2 4
PAGE 169
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Solution: The player A wishes to obtain the largest possible value vij by
choosing one of his strategies I, II, III, while the player B is determined
to make A gain the minimum possible by choice of his strategies I, II,
III. The player A is called maximizing player and player B is called
minimizing player.
Find the smallest value from each row to get row minimum for player A
and also find the greatest value from each column to get column maxi-
mum for player B.
In first, second and third row the smallest values are 3, 0, 4 respec-
tively and in first, second and third column the greatest values are 5, 0,
6 respectively. Therefore we have,
Player A should try to maximize his least gains or to find maxi [minj
(vij)]. The best choice for player A is to maximize his least gains by
selecting his II strategy which assures at most the gain 0. On the other
hand, player B can argue to keep A’s gain the minimum or to find minj
[maxi (vij)]. Player B then selects the strategy II that minimizes his max-
imum losses to 0.
170 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Player A’s selection is called maximum strategy and his corresponding Notes
gain is called maximin value or lower value of the game and Player B’s
selection is called minimax strategy and his corresponding loss is called
minimax value or upper value of the game.
In this case minimax and maximin values are equal, the corresponding
pure strategies for player A is II and for player B is II, are called optimal
strategies and the game is said to have a saddle point.
Example: Consider the following game and find the saddle point.
Player B
I II III
I 3 4 8
Player A
II 8 5 6
III 6 7 6
PAGE 171
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Example: In a game of matching coins with two players, suppose one player
wins Rs. 2 when there are two heads and wins nothing when there are two
tails; and losses Rs. 1 when there are one head and one tail. Determine the
payoff matrix, the optimal strategy for each player and the value of the game.
Solution: According to the given information, the payoff matrix of player
A is given by,
Player B
H T
Player A H 2 0
T 0 1
172 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Here in this case the maximin value = 1 and minimax value = 2, so Notes
the matrix is without saddle point.
Now to find the optimal strategy, let us consider that player A plays H
(head) with probability x and T (tails) with probability (1 x), then if
player B plays H all the time, A’s expected gain will be
E(A, H) = x.2 + (1 x).(1) = 3x 1
If player B plays T all the time, A’s expected gain will be
E(A, T) = x.(1) + (1 x).(0) = x
So, if the A chooses x such that, E(A, H) = E(A, T) = E(A), then this
will determine the optimal strategy for A.
So, 3x 1 = x
Or, x = 1/4
And 1 x = 1 (1/4) = 3/4
Therefore, the optimal strategy for player A is to play H and T with
probability 1/4 and 3/4 respectively. Therefore, the mixed strategy for
player A is (1/4, 3/4) and the expected gain for A is,
E(A) = (1/4).2 + (3/4).(1) = 1/4
Similarly, let us consider that player B plays H (head) with probability
y and T (tails) with probability (1 y), then if player A plays H all the
time, B’s expected gain will be
E(B, H) = y.2 + (1 y).(1) = 3y 1
If player A plays T all the time, B’s expected gain will be
E(B, T) = y.(1) + (1 y).(0) = y
So, if the B chooses x such that, E(B, H) = E(B, T) = E(B), then this
will determine the optimal strategy for B.
So, 3y 1 = y
Or, y = 1/4
And 1 y = 1 (1/4) = 3/4
Therefore, the optimal strategy for player B is to play H and T with
probability 1/4 and 3/4 respectively. Therefore, the mixed strategy for
player B is (1/4, 3/4).
and E(B) = (1/4).2 + (3/4).(1) = 1/4
PAGE 173
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Here, E(A) = E(B) = 1/4, the complete solution of the game is,
Player A should play H and T with probabilities 1/4 and 3/4
respectively. Thus A’s optimal strategy is x0 = (1/4, 3/4).
Player B should play H and T with probabilities 1/4 and 3/4
respectively. Thus B’s optimal strategy is y0 = (1/4, 3/4).
The expected value of the game is 1/4 to player A. Hence (x0, y0)
is the strategic saddle point of this game.
Player B
y1 y2
Then the optimal strategies (x1, x2) for player A & (y1, y2) for player B
are determined by,
x1/x2 = (v22 v21)/(v11 v12)
and y1/y2 = (v22 v12)/(v11 v21)
and the value of the game v to the player A is given by,
Player B
y1 y2
174 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Then the equations for strategies of player A and player B are given by, Notes
x1v11 + x2v21 = v (1) y1v11 + y2v12 = v (4)
x1v12 + x2v22 = v (2) y1v21 + y2v22 = v (5)
x1 + x2 = v (3) y1 + y2 = v (6)
Now subtracting equation (2) from equation (1), we get
x1 (v11 v12) + x2 (v21 v22) = 0 (7)
Multiplying equation (3) by (v11 v12), we get
x1 (v11 v12) + x2 (v11 v12) = (v11 v12) (8)
Solving equations (7) and (8), we get
x2 = (v11 v12)/[(v11 + v22) (v12 + v21 @
Using value of x2 in equation (3), we get
x1 = (v22 v21)/[(v11+ v22) (v12 + v21)] (10)
'LYLGLQJ HTXDWLRQ E\ HTXDWLRQ ZH JHW
x1/x2 = (v22 v21)/(v11 v12)
Similarly, subtracting equation (5) from equation (4), we get
y1 (v11 v21) + y2 (v12 v22) = 0 (11)
Multiplying equation (6) by (v11 v21), we get
y1 (v11 v21) + y2 (v11 v21) = (v11 v21) (12)
Solving equations (11) and (12), we get
y2 = (v11 v21)/[(v11 + v22) (v12 + v21)] (13)
Using value of y2 in equation (6), we get
y1 = (v22 v12)/[(v11 + v22) (v12 + v21)] (14)
Dividing equation (14) by equation (13), we get
y1/y1 = (v22 v12)/(v11 v21)
Now using values of x1 and x2 in equation (1), we get
v = (v11v22 v12v21)/[(v11 + v22) (v12 + v21)]
PAGE 175
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes
7.9.1 Steps to Solve 2 u 2 Games without Saddle Point
The following are the steps to solve 2 u 2 games without saddle point
arithmetically:
Find the difference of two numbers in column I, and put it under
column II (Neglect the negative sign, if exist).
Find the difference of two numbers in column II, and put it under
column I (Neglect the negative sign, if exist).
Repeat the above two steps for two rows also.
The values thus obtained are called the oddments, which are the frequencies
the players must use their courses of action in their optimum strategies.
Example: Two players A and B, without showing each other, put on a
table a coin, with head or tail up. Player A wins Rs. 8 when both the
coins shows head and Rs. 1 when both are tails. Player B wins Rs. 3
when the coins do not match. Given the choice of being matching play-
er A or non-matching player B, which one would you choose and what
would be your strategy?
Solution: According to the given information, the payoff matrix for
player A is given by,
Player B
H T
Player A H 8 3
T 3 1
Here in this case the maximin value = 3 and minimax value = 1, so the
matrix is without saddle point and the optimal strategies will be mixed
strategies.
Step 1: Difference of two numbers in column I is 8(3) = 11, put it
under column II.
Step 2: Difference of two numbers in column II is 31 = 4, and put
it under column I (Neglect the negative sign).
Step 3: Repeat the above two steps for two rows also.
176 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 177
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes dominance, the payoff matrix can be resized by deleting those strategies
which are dominated by the others. The concept of dominance is espe-
cially useful for the evaluation of two-person zero-sum games where a
saddle point does not exist.
The dominance properties are as follows:
If all the elements of a row say ith, are less than or equal to the
corresponding elements of any other row, say rth row, then ith row
is dominated by the rth row.
If all the elements of column, say jth, are greater than or equal to
the corresponding elements of any other column, say sth then jth
column is dominated by the sth column.
If Row I is dominated by the average of Row II and Row III then
Row I will be removed from the matrix. For example, if Row
, DYHUDJH RI 5RZ ,, DQG 5RZ ,,, WKHQ 5RZ , ZLOO EH UHPRYHG
If Column I is dominated by the average of Column II and Column
III then Column I will be removed from the matrix. For example,
LI &ROXPQ , DYHUDJH RI &ROXPQ ,, DQG &ROXPQ ,,, WKHQ &ROXPQ
I will be removed.
Let M = [vij] be the payoff matrix of an m x n rectangular game. If the
ith row of M is dominated by rth row of M, then the deletion of ith row of
M does not change the set of optimal strategies of player A. Further, if
the jth column of M is dominated by sth column of M, then deletion of jth
column of M does not change the set of optimal strategies of player B.
Example: Use dominance rule to solve the following rectangular game
whose payoff matrix to A is given by,
Player B
I II III IV V VI
I 0 0 0 0 0 0
II 4 2 0 2 1 1
Player A III 4 3 1 3 2 2
IV 4 3 7 5 1 2
V 4 3 4 1 2 2
VI 4 3 3 2 2 2
178 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Solution: In the given payoff matrix, the all the elements of I and II Notes
rows are less than or equal to the III row, therefore I and II rows will
be dominated by row III. Hence the reduced matrix is,
Player B
I II III IV V VI
III 4 3 1 3 2 2
Player A IV 4 3 7 5 1 2
V 4 3 4 1 2 2
VI 4 3 3 2 2 2
All the elements of I, II and VI columns are greater than or equal to the
V column, therefore I, II and VI columns will be dominated by column
V. Hence the reduced matrix is,
Player B
III IV V
III 1 3 2
Player A IV 7 5 1
V 4 1 2
VI 3 2 2
All the elements of VI row are less than or equal to the V row, therefore
VI row will be dominated by row V. Hence the reduced matrix is,
Player B
III IV V
III 1 3 2
Player A
IV 7 5 1
V 4 1 2
The V column is superior to the average of III and IV columns; therefore
V column will be dominated. Hence the reduced matrix is,
Player B
III IV
III 1 3
Player A
IV 7 5
V 4 1
PAGE 179
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes The average of III and IV rows is greater than V row; therefore V row
will be dominated. Hence the reduced matrix is,
Player B
III (y3) IV (y4)
Now, for 2 u 2 game without saddle point, solve the following simulta-
neous equations:
For player A:
1.x3 + 7.x4 Y[3 5.x4 Y[3 + x4 = 1
Solving these equations we have,
x3 [4 Y
For player B:
1.y3 +3.y4 Y\3 5.y4 Y\3 + y4 = 1
Solving these equations we have,
y3 \4 Y
Therefore the complete solution is,
Optimal strategy for player A (0, 0, 6/7, 1/7, 0, 0)
Optimal strategy for player B (0, 0, 4/7, 3/7, 0, 0)
The value of the game to player A = 13/7
The value of the game to player B = 13/7
180 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Example: Find the solution of game theory problem (6 u 2 game) using Notes
graphical method:
Player B
Strategies B1 B2
A1 1 3
Player A A2 3 5
A3 1 6
A4 4 1
A5 2 2
A6 5 0
PAGE 181
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Notes Step 3: Find the feasible region and the point of optimal solution.
Find the feasible region and the point of optimal solution for player B,
as player B has only 2 strategies.
Since player B is a minimax player therefore, consider the lowest point
(P) of the upper feasible region (upper boundary) of all the lines in the
graph.
Step 4: Find the optimal strategies and the value of the game.
The point of optimal solution (P) occurs at the intersection of two lines
corresponding to A2 and A4 strategies of player A, therefore the optimum
strategies for player B can be calculated by the following equations:
3y + 5(1 \ Y\ y) = v
Solving these equations, we get
\ y = 1 Y
Similarly, Let for the strategies A2 and A4 of player A, the respective
probabilities are x and (1 x).
Therefore the optimum strategies for player A can be calculated by the
following equations:
3x + 4(1 [ Y[ x) = v
Solving these equations, we get
[ x = 1 Y
Therefore the solution is,
Therefore the optimum strategies for player A (0, 3/5, 0, 2/5, 0, 0)
Therefore the optimum strategies for player B (4/5, 1/5)
The value of the game is 17/5.
Example: Find the solution of game theory problem (2 u 3 game) using
graphical method:
Player B
Strategies I II III
Player A I 1 3 11
II 8 5 2
182 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
Solution: Here, in this case, Minimax value = 11 and Maximin value = 2 Notes
Therefore, the game has no saddle point.
Step 1: Draw two parallel lines 1 unit distance apart and mark a scale on each.
Since player A has only two strategies draw the graph for player A. The
two parallel lines represent strategies I and II of player A with probabil-
ities x and (1 x) respectively.
Step 2: Draw the lines corresponding to each strategy of player B and
both the strategies of player A.
Step 3: Find the feasible region and the point of optimal solution for
player A, as player A has only 2 strategies.
Since player A is a maximin player therefore, consider the highest point (P)
of the lower feasible region (lower boundary) of all the lines in the graph.
Step 4: Find the optimal strategies and the value of the game.
The point of optimal solution (P) occurs at the intersection of two lines
corresponding to II and III strategies of player B, therefore the optimum
strategies for player A can be calculated by the following equations:
3x + 5(1 [ Y[ x) = v
PAGE 183
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
7.14 Summary
In the theoretical discipline of social sciences known as “game theory,”
results of a certain scenario are predicted using a mathematical model.
Those who work in political science, business, or poker frequently
utilise it to forecast possible outcomes for circumstances relevant to
their disciplines. In order to forecast how individuals or organizations
will behave, game theory uses sequential games to mimic a number
of real-world, strategic situations. Yet, the optimal course of action is
typically to cooperate to secure the most favourable, symmetric outcome
for all parties. The dominating strategy is frequently for a player to
choose the option that benefits them the most.
When examining a particular economic environment, economists can use
game theory to forecast the actions that businesses (or players) will take.
Private companies may also use it to do business or to proactively track
and assess the various elements and competitive practises inside their
target economy. Instructors in business schools may also employ game
theory simulations to teach their students to a variety of techniques and
problem-solving ideas that they may observe applied in the real world.
Notes 2. Define and discuss the various types of games classified based on
information availability, such as games with perfect information and
games with imperfect information. Explain how players’ knowledge
about the game’s structure and their opponents’ actions influences
their decision-making processes.
3. Describe the concept of a saddle point in game theory. Explain how
saddle points, optimal strategies, and the value of the game are
determined in matrix games. Provide a step-by-step explanation of
the process involved in identifying a saddle point and determining
optimal strategies for players.
4. Discuss the minimax (maximin) principle in game theory. Explain
how this principle is applied in decision-making under uncertainty
and competition. Provide examples to illustrate how players use
the minimax principle to minimize potential losses and maximize
gains in zero-sum games.
5. Explain the dominance rule in game theory. Discuss how dominant
strategies and dominated strategies are identified in games, and how
they affect players’ decision-making processes. Provide examples and
discuss the implications of the dominance rule in real-life strategic
interactions and competitive scenarios.
7.17 References
Maynard Smith J. Evolution and the Theory of Games. Cambridge,
8. &DPEULGJH 8QLYHUVLW\ 3UHVV
Von Neumann J., & Morgenstern O. (2004). Theory of Games and
Economic Behaviour. 4th ed., Princeton, NJ: Princeton University
Press
186 PAGE
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
PAGE 187
Department of Distance & Continuing Education, Campus of Open Learning,
School of Open Learning, University of Delhi
QUANTITATIVE TECHNIQUES
Department of Distance and Continuing Education DEPARTMENT OF DISTANCE AND CONTINUING EDUCATION
University of Delhi UNIVERSITY OF DELHI
AS PER THE UGCF-2022 AND NATIONAL EDUCATION POLICY 2020