Chapter Two: Matrix Algebra and its Applications 1

CHAPTER TWO

Matrix Algebra and its applications

Algebra - is a part of mathematics that deals with operations (+, -, x÷).

Matrix is a rectangular array of real numbers arranged in m rows and n columns.
Like sets, it is symbolized by a bold face capital letter enclosed by brackets or
parentheses as:
a

a 11 a
12


1n

A  a a 21 a
22  in which aij
2n
 
 a a m1 
m2      a  mn

are real nos.

Each number appearing in the array is said to be an element or component, of the

matrix. Elements of a matrix are designated using a lowercase form of the same
letter used to symbolize the matrix itself. These letters are subscript, as aij, to give
the row and column location of the element within the array. The first subscript
always refers to the row location of the element; the second subscript always refers
to its column location. Thus, component aij is the component located at the
intersection of the ith row and the jth column.

The number of rows, m, and the number of columns, n, of the array give its order,
or its dimensions, mxn (read “m by n”) = a mxn or [a ij] (mxn).
Example: the following are examples of matrices

1 7 
 
 
A  5

4
3 

2
 This is a 3 x 2 matrix
 

ELEMENT
a12= 7
a21 = 5
a32 = 2
a23 = X - Because is a 3 x 2 matrix.
1 5 9 15
2 6 10 20
  This is a 4 x 4 matrix Elements X = 45 x32 = 7
3 7 11 30 44
 
4 8 12 45

IMPORTANCE OF MATRICES
Matrices provide a most convenient vehicle for organizing and storing large
quantities of data. Because the basic idea is to organize the data, we cannot over
emphasize the importance of the location of each number with in the matrix. It is
not simply a matter of putting numbers in to rows and columns; each row-column
location with in each matrix carries with it special interpretation; a matrix is, in

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 2

essence, a tool for organizing vast quantities of data. Matrices are used to represent
complex systems and operations by compact entities.
Matrix representations are possible
- Transportation matrix
- Distance matrix
- Cost matrix
- Brand switching

TYPES OF MATRICES
1. VECTOR MATRIX - is a matrix which consists of either one row or one column.
That is, it is an mx1 or a 1 x n matrix.
1.1. Row vector = is a 1 x n matrix
E.g. W = [-1, 0, 6]
1.2 Column Vector - is a mx1 matrix
 2
5 

E.g. B =  
 7
 
 0

The transpose of an mxn matrix denoted A -t is an nxm matrix whose rows are the
columns in A (in the same order) and whose columns are the rows in A (in the same
order).
1 1 
2 3 10  
4 7

  2 
  -t 
5 8


If A  4

7
5

8
6

9
11 

12
 then A = A  
3

10
6

11
9



12 
   

t
Note that a = aij ij

The transpose of a row vector is a column vector and the transpose of a column
vector is a row vector.

2. Square Matrix - is a matrix that has the same number of rows and columns. It is also
called an nth order matrix.
1 0
A   
E.g. 2x2, 0

1

.
 

3. NULL (ZERO) MATRIX - is a matrix that has zero for every entry. It is generally
denoted by Omn. In matrix operations it is used in much the same way that the
number zero is used in regular algebra. Thus, the sum of a zero matrix and any
matrix gives that given matrix and the product of a zero matrix and any matrix
equals a zero matrix
4. IDENTITY MATRIX - a square matrix in which all of the primary diagonal entries
are ones and all of the off diagonal entries are zeros. Generally it is denoted as I n.
Primary diagonal represents: a11, a22, a33, a44, --- ann entries.

1 0 1
0
0 0 0
0
A    1 0
I2 = 0

1

, I4 = A  
0

0 1

0

  0 0 0 1

The product of any given matrix and the identity matrix is the given matrix it self.
That is, A x I = A and I.A = A. Thus, the identity matrix behaves in matrix
multiplication like the number 1 in an ordinary arithmetic.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 3

5. SCALAR MATRIX - is a square matrix where elements on the primary diagonal are
the same and the rest zeros.
NB: An Identity matrix is a scalar matrix, but a scalar matrix may not be an identity
matrix
6. DIAGONAL MATRIX- a square matrix where elements on the primary diagonal are
consecutive and others zeros.
7. EQUAL MATRICES -Two matrices A & B, are said to be equal only if they are of
the same dimensions and if each element in A is identical to its corresponding
element in B; that is, if and only if a ij = bij for every pair of subscripts i and j. If A =
B, then B = A; or if A≠B, then B ≠A.

1 2 1 2
A     
3

4

Is equal to B = 3

4

   

1 2 4 2
A     
However; 3

4

is not equal to C = 3

1

   

Even though they contain the same set of numerical values, A and C are not equal
because their corresponding elements are not equal; that is, a 11 ≠ C11 and so on.

MATRIX OPERATIONS (ADDITION, SUBTRACTION, and

MULTIPLICATION)

Matrix Addition (subtraction)

Two matrices of the same dimensions are said to CONFORMABLE FOR
ADDITION. The addition is performed by adding corresponding elements from the
two matrices and entering the reset in the same row-column position of a new
matrix [element-wise addition].

If A and B are two matrices, each of size mxn, then the SUM of A and B is the mxn
matrix C whose elements are:
Cij = Aij + bij for i = 1, 2, ------- m
j = 1, 2, -------- n.

Laws of Matrix Addition

The operation of adding two matrices that are conformable for addition has these
two basic properties:
1. A + B = B + A ---- The commutative law of matrix addition.
2. (A+B) +C = A+ (B+C) -------- the associative law of matrix addition.

1 3 7 9  7 9 
eg      
2

4

+ 8

 10

= 10

 6

     

Given that two matrices do have the same dimension, the way we subtract a matrix
from another matrix is the same as the way we add two matrices.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 4

Matrix Multiplication

A. Matrix Multiplication by a Constant (Scalar Multiplication)

A matrix can be multiplied by a constant by multiplying each component in the
matrix by a constant. The result is a new matrix of the same dimensions as the
original matrix.

If K is any real number and A is an mxn matrix, then the product KA is diffident to
be the matrix whose components are given by k times the corresponding component
of A; that is,
KA= [Kaij] (mxn).

E.g. If X = [6 5 7], then 2X = [(2x6) (2x5) (2x7)]

2X = [12 10 14]
Laws of Scalar Multiplication
The operation of multiplying a matrix by a constant (a SCALAR) has the following
basic properties. If x and y are real numbers and A and B are mxn matrices,
conformable for addition, then:
1. XA = AX
2. (X+Y)A = XA+YA
3. X (A+B) = XA + XB
4. X (YA) = XY (A)

B. Vector-by-Vector multiplication
In multiplying two vectors always a row vector is written in the first position and
the column vector in the second position. Each component of a row vector is
multiplied by the corresponding component of the column vector to obtain a result
known as PARTIAL PRODUCT. The sum of all partial products is called
INNER/DOT PRODUCT of two vectors, and this is a number not a vector. In other
words, Vector by Vector results in a real number rather than a matrix.

E.g. Consider the product (AB) of the following row and column vectors.
 2

A 3 6
5 
4 2 ’ B=  
 7
 
 0

3x2=6
4 x 5 = 20 partial products
-2 x 7 = -14

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 5

6x0=0
12 Inner/Dot Product

C. Matrix by Matrix Multiplication

If A and B are two matrices, the product AB is defined if and any if the number of
columns in A is equal to the number of rows in B, i.e., if A is an m x n matrix, B
should be an n x b. If this requirement is met, A is said to be CONFORMABLE to B
FOR MULTIPLICATION. The matrix resulting from the multiplication has
dimensions equivalent to the number of rows in A and the number of columns in B.

Matrix by matrix multiplication indicates a row by column multiplication, where

the entry in the ith row and jth column of the product AB is obtained by multiplying
the entries in the i th row of A by the corresponding entries in the j th column of B and
then adding the results. That is, to obtain the entry in the i th row and jth column of
the product AB, use the ith raw of A and the jth column of B in the following form:

The first element in the raw is multiplied by the first element in the column; the
second element in the row is multiplied by the second element in the column and so
on until the nth row element is multiplied by nth column element. These products
are then summed up to obtain the single number that is the product of the two
vectors.

If A is a matrix of dimension n x m (which has m columns) and B is a matrix of

dimensions p x q (which has p rows) and it m is different from p, the product AB is
not defined. That is, multiplication of matrices is possible only if the number of
columns of the first equals the number of rows of the second.

If A is of dimension n x m and if B is of dimension m x p, then the product A.B is of

dimension n x p.

Example
2 3 4  1 7
A    0 
6

9 7

B= 

8 

  5 1 

A.B = (2x-1) + (3x0) + (4x5) (2x7) + (3x8) + (4x1)

= 18 42

= (6x-1) + (9x0) + (7x5) (6x7) + (9x8) + (7x1)

= 29 121

18 42 
AB =  29
 121


Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 6

AB = (-1x2) + (7x6) = 40
 40 60 45
 48 56 
(0x3) + (8x9) = 72 AB = 

72


16 24 27 

Special Properties of Matrix Multiplication

1. The Associative and distributive laws of ordinary algebra apply to matrix

multiplication. Given three matrices A, B and C, which are conformable for
multiplication,
 A (BC) = (AB) C -------------------- Associative law, not C (AB).
 A (B+C) = AB + AC -------------- Distributive law
 (A+B) C = AC + AB -------------- Distributive law

2. The commutative law of multiplication does not apply to matrix multiplication. For
any two real numbers X and Y, the product XY is always identical to the product
YX. But for two matrices A and B, it is not generally true that AB equals BA. (In the
product AB, we say that B is pre multiplied by A and that A is post multiplied by
B). In many instances for two matrices A and B, the product AB may be defined
while the product BA is not defined, or vice versa.

In some special cases, AB does equal BA. In such special cases A and B are said to
Commute.

3. The product of two matrices can be the zero matrix even though neither of the two
matrices them selves is zero matrix! We cannot conclude from the result AB = 0 that
at least one of the matrices A or B is a zero matrix.
3 0 0 0 0 0 
2 0 7 4
A = 

0 
0 
, B = 

10 

, AB =
1 0  8 3 2 
0 0 0
0 0 0
 

0 0 0


4. We cannot, in matrix Algebra, necessarily conclude from the results AB = AC that B

= C, even if matrix A is not equal to a zero matrix. Thus the CANCELLATION LAW
does not hold, in general, in matrix multiplication.

1 3  4  1 1 2 
A    3 
 2

 6

,B 2

5 

,C 2

4

     

10 14 
 
AB = AC =  20

 28

but B ≠ C.
 

The Multiplicative Inverse of a Matrix

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 7

If A is a square matrix of order n, then a square matrix of its inverse (A -1) of the
same order n is said to be the inverse of A, if and only if AA -1 = I = A-1A.
Two square matrices are inverse of each other if their product is the identity matrix:
I = AA-1 = A-1A.

Not all matrices have an inverse. In order for a matrix to have an inverse, the matrix
must, first of all, be a square matrix. Still not all square matrices have inverse. If a
matrix has an inverse, it is said to be INEVITABLE or NON-SINGULAR. A matrix
that doesn’t have an inverse is said to be SINGULAR. An inevitable matrix will
have only one inverse; that is, it a matrix does have an inverse, and that inverse is
unique.
In short:
 Inverse of a matrix is defined only for square matrices
 If B is an inverse of A, then A is also an inverse of B.
 Inverse of a matrix is unique.
 If matrix A has an inverse, A is said to be inevitable and not all square
matrices are inevitable.
1 1
eg  
E.g. 1

1

 

Finding the Inverse of a Matrix

Let us begin by considering a tabular format where the square matrix. A is

augmented with an identity matrix of the same order, as [A/I]. This process is called
ADJOINING.
Now, if the inverse matrix A -1 were known, we could multiply the matrices on each
side of the vertical line by A-1, as [AA-1/A-1I]

Then, because AA-1 = I and A-1 I = A-1, we would have [I/A-1]. We do not follow this
procedure, because the inverse is not known at this juncture; we are trying to
determine the inverse. We instead employee a set of permissible row operations on
the augmented matrix [A/I] to transform A on the left side of the vertical line in to
an identity matrix (I). As the identity matrix is formed on the left of the vertical line,
the inverse of A is formed on the right side. The allowable manipulations are called
ELEMENTARY ROW OPERATIONS. These Elementary Row Operations are
operations permitted on the row of a matrix.

In a matrix Algebra there are 3 types of row operations.

i. Any pair of row in a matrix may be interchanged /Exchange operations/.
Interchanging rows.
ii. A row can be multiplied by any non-zero real number /Multiple
operations/. The multiplication of any row by a non-zero number.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 8

iii. A multiple of any row can be added to any other row /Add-A-Multiple
operations/. The addition /subtraction of (a multiple of) one row
to/from) another row.

4 3 2   2 6 7
A  4 
E.g. 1.  2

6 7

, B 

3 2


= interchanging
   

rows

4 3 2  8 6 4
A  A 
2.  2

6 7

B=  2

6 7

= multiplying the first row
   

by 2.

4 3 2  4 3 2 
A   
3.  2

6 7

B= 6

12 11

= Multiplying the first row by 2
   

and add to 2nd row.

Theorem on row operations

A row operation performed on product of two matrices is equivalent to row
operation performed on the pre-factor.

Consider the following AB = C

 1 3 1 2 
2 9 13 
A  1 
2

3 4 

B= 
2
1
3


C, = 13
 19

   

Interchange R1 with R2

2 4 1 2 
3 13 19
A  1 
1

2 3

B= 
2
1
3


C, = 9
 13 

   

Basic Procedures to Find the Inverse of a Square Matrix

1. To get ones first in a column and next zeros (within a given column)
2. To get zeros first in a matrix and next ones.

Ones First: Try to set ones first in a column and then zeros of the same column. G0
from left to right
Zeros First: Find the off diagonal zeros first, and following this obtain ones on the
main diagonal. It can simplify the work involved in hand calculation by avoiding
fractions until the last step.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 9

MATRIX APPLICATIONS

Solving Systems of Linear Equations

1. n by n systems
Systems of linear equations can be solved using different methods. Some are:
 Elimination method for 2 variable problems (equations).
 Matrix method
i. Inverse method
ii. Cramer’s rule – using determinants (independent study)
iii. Gaussian Method.

Inverse Method

To solve systems of linear equations using the inverse method the coefficient matrix
should be inevitable, and it involves the following steps:
1. Put all equations in a matrix form (square matrix form).
2. Find the inverse of the coefficient matrix.
3. Multiply the inverse with right hand side values (vector of constants)

2. X+Y = 2
2x + 2y 4

The inverse method provides us with unique solution, or no solution and infinite
solution (with out separating them).

Gaussian Method: developed by Karl F. Gauss (1771-1855)

Solving systems of linear equations using the Gaussian method involves the
following steps:
1. Write all equations in a matrix form.
2. Change coefficient matrix in to identity matrix and apply the same
commentary row operations on the vector of constants
3. The resulting value (of the RHS vector) will be the solution.

Ax = B
Ix = C
x=C
The Gaussian Method helps us to obtain:
 Unique solution
 No. Solution
 Infinite solution

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 10

E.g. 1. 2x + 3y = 4 2. x + y = 2 3. x + y = 5
x + 2y = 2 2x + 2Y = 4 x+y=8

2 3   4

 1

2
 
 2
IX = c
X=C

Therefore, Gaussian method makes a distinction between no solution and infinite

solution, unlike the inverse method.

Summarizing our results for solving an “n” by “n” system, we start with matrix
(A/B), and attempt to transform it in to the matrix (I/C).
One of the three things will result:

1. An n by n matrix with the unique solution; e.g.

1 0 0 10
 
0 1 0 5
0 0 1 3 


2. A row that is all zeros except in the constant column, indicating that there are no
solutions; e.g.

1 0 0 3
 
0 1 0 5
0 0 0 7


3. A matrix in a form different from (1) and (2), indicating that there are an unlimited
number of solutions. Note that for an n by n system, this case occurs when there is a
row with all zeros, including the constant column; e.g.

1 0 2 5
 
0 1 3 3
0 0 0 0 


2. M by n linear systems
The mxn linear systems are those systems where the number of rows (m) and
number of columns (n) are unequal or it is the case where the number of equations
(m) and the number of variables (n) are unequal. And it may appear as m>n or m<n.
2.1 Linear equations where m>n
To solve an m by n systems of equations with m>n, we start with the matrix (A/B),
and attempt to transform it in to the matrix (I/C). One of the three things will result:

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 11

1. An n by n identity matrix above m-n bottom rows that are all zeros, giving the
unique solution
E.g.

2. A row that all zeros except in the constant column, indicating that there are no
solutions.
E.g.

3. A matrix in a form different from (a) and (2), indicating that there are an
unlimited number of solutions.

E.g.

2.2. Linear equations where m<n

Our attempt transform (A/B) in to (I/C) in the case where m<n will result in:
1. A row that is all zeros except in the constant column, indicating that there are no
solutions. Or

2. A matrix in a form different from (2), indicating that there are an unlimited
number of solutions.

“EVERY SYSTEM OF LINEAR EQUATIONS HAS NO SOLUTION, EXACTLY ONE

SOLUTION OR INFINITELY MANY SOLUTIONS.”

WORD PROBLEMS

Steps

1. Represent one of the unknown quantities by a letter usually x and express other
unknown quantities if there is any in terms of the same letter.
2. Translate the quantities from the statement of the problem in to algebraic form
and set up an equation.
3. Solve the equation (equations) for the unknown that is represented by the letter
and find other unknowns from the solution.
4. Check the findings according to the statement in the problem.

Example.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 12

1. The weights (in pounds) of six people before taking a weight reduction program
were 350, 249, 260, 195, 275, and 295. The weights of these same people after the
weight reduction program are 345, 200, 220, 140, 200, and 230, respectively.
Summarize this information in a (6 by 2) matrix.

2. A manufacturing firm which manufactures office furniture finds that it has the
following variable costs in dollars.
Desks Chairs Tables Cabinets
50 20 15 25 
30 15 12 15 
 
30
 15 8 20 


Assume that an order of 5 desks, 6 chairs, 4 tables and 12 cabinets has just been
received. What are the total material, labor and overhead costs associated with the
production of ordered items? Answer: Birr 1,710.

3. Olympus manufacturing company produces two types of boats: one-person and

two-person models. The company has two plants x and y at different parts of the
country. In both plants, there are two departments, fabricating and finishing. A one-
person boat requires 4-labor hr in the fabricating department and 1 labor hr in the
finishing department. The two-person boat requires 6-labor hr in the fabricating
department and 1.5-labor hr in the finishing department. Suppose the hourly rates
of labor cost in the fabricating and finishing departments be Br 8 and Br 6
respectively at plant x, and Br 7 and Br 4 at plant Y. Using matrix algebra, find the
labor cost of making one unit of each product at each of the two plants. Interpret the
results.
Answer:
Product
Plant One person Two person

X 38 57

Y 32 48

4. Asrat Carpet Company has in inventory 1,500 square yards of wool and 1,800
square yards of nylon for the manufacture of carpeting. Two grades of carpeting are
produced. Each roll of superior grade carpeting requires 20 sq. yards of wool and 40
square yards of nylon. Each roll of quality-grade carpeting requires 30 square yards
of wool and 30 square yard of nylon. If Asrat would like to use all the material in
inventory, how many rolls of superior and how many rows of quality carpeting
should be manufactured? Answer: 15 and 40.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 13

5. A manufacturer is costing out one product line which consists of three different
models, A, B, and C. These models are assembled from three types of parts, 1,2 and
3. The manufacturer would like to produce such quantities of the three models as to
completely deplete the inventory of parts of hand, in the final production run. Each
model a uses one unit of part 1, three units of part 2, and two units of part 3. Each
unit of model B uses one unit of part 1, two units of part 2, and one unit of part 3.
Each model C uses two units of part 1, and three units of part 3. Inventory records
show that there are on hand 1,500 units of part 1 and 1,900 units each of part 2 and
3. How many of each model should the manufacturer plan to produce? Answer:
100, 800, and 300.

6. Alemayehu invested a total of 10,000 in three different savings accounts. The

accounts paid simple interest at an annual rate of 8 percent, 9 percent and 7.5
percent respectively. Total interest earned for the year was Br 845. The amount in
the 9 percent account was twice the amount invested in the 7.5 percent account.
How much did Alemayehu invest in each account? Br 1,000, Br 6,000 and Br 3,000

7. A person invests in A, B and C rated bonds. The average yield is 8% on A bonds,

6% on B bonds, and 7% on C bonds. Twice as much is invested in C bonds as B
bonds. Moreover, the total annual return for all three types of bonds in Br. 2800.
How much is invested in each type of bond if the total investment is
a. 37,500? Answer: Br 22,500, 5,000 and 10,000
b. 40,000? Br 10,000, 10,000 and 20,000
8. A certain manufacturer produces two products, P and Q. Each unit of product P
requires in its production 20 units of raw material A and 10 units of raw material B.
Each unit of product Q in its production requires 30 units of raw material A and 50
units of raw material B. There is a limited supply of 1200 units of raw material A
and 950 units of raw material B. how many units of product P and Q can be
produced if we are to exhaust the supply of raw materials? Answer: 45 and 10.

9. Attendance records indicate that 40, 000 people attended the 12 th African Youth
championship at its opening ceremony at the Addis Ababa Stadium. Total ticket
receipts were Br 1, 750,000. Admission prices were Br 37.5 for the second class and
Br 62.5 for the first class. Determine the number of people who attended the opening
ceremony at first class and second class. Answer: 30,000 and 10,000.

10. A mixture containing X pounds of ingredient A, Y ponds of ingredient B and Z

pounds of ingredient C is to be made. The mixture is expected to have a weight of 5
pounds and contain 1500 units of vitamin and 2,500 units of calories. The vitamin
and caloric content of the three ingredients is given below.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 14

Determine how many pounds of each ingredient should in the 5-pound mixture.
Answer: Unlimited solution.

11. Two departments of a firm A and B need different amounts of the same product. The
following table gives the amounts of the products needed by the department.

Two suppliers, Abebe and Tefera supply these three products with the unit-selling
price given below:

Abebe Tefera

Steel 300 280

Plastic 150 100

Wood 150 200

Required:
a) Use matrix multiplication to find how much these two orders will cost at the
two suppliers.
b) From which supplier should each department make its purchase? Answer:
Department A from Tefera, and Department B from Abebe.

MARKOV CHAINS

Concept, Model and Solutions

This model is a forecasting model. It is probabilistic/ stochastic model. A Russian

mathematician called Andrew Markov around 1907 developed this model.

Markov chains are models which are useful in studying the evolution of certain
system over repeated trials. These repeated trails are often successive time periods
where the state (out come, condition) of the system in any particular time period
cannot be determined with certainty. Therefore, a set of transition probabilities is
used to describe the manner in which the system makes transition from one period
to the next. Hence, we can predict the probability of the system being in a particular
state at a given time period. We can also talk about the long run/equilibrium, steady
state.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 15

System - which we want to study, machine, and person

Trials - successive time period any convenient length of time day, week, month,
year, etc.
State/out come, condition - the system can have various number of out comes.
Transition probabilities - set of input data, and are assumed to be constant.
Long/stead state - the system cannot change any more. There is the same probability
between n and n + 1 period after the long period.

The necessary assumptions of the chain are:

1. The system has a finite number of states - the out comes of the system should be
finite.
2. The system condition/outcome, state in any given period depends on its state in
the preceding period and on the transition probabilities
3. The transition probabilities are constant over time.
4. Changes in the system will occur once and only once each period.
5. The transition period occurs with regularities.
6. The states are both mutually exclusive and collectively exhaustive.
7. The system is a closed one, that is, there will be no arrival or exits from the
system.

Information flow in the Analysis

The Markov model is based on two sets of input data

 The set of transition probabilities.
 The existing or initial or current conditions or states.

The Markov process, therefore, describes the movement of a system from a certain
state in the current state/ time period to one of n possible states in the next stage.
The system move in an uncertain environment all that is known is the probability
associated with any possible move or transition. This probability is known as
transition probability symbolized by P ij. It is the likelihood that the system which
is currently in state i will Smoke to state j in the next period.

From these inputs the model makes two predictions usually expressed as vectors:
1. The probabilities of the system being in any state at any given future time period.
2. The long run / equilibrium, steady state probabilities.

The set of transition probabilities are necessary for both predictions (time period n,
and steady state), but the initial state is needed for only the first prediction.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 16

In put data Predictions/ outcomes

Set of transition
Probabilities Steady states/ long run states
About past

Current/initial state the probability of the system

being in any state at any given time
About today

Markov chain analysis used among other things in Market share Analysis. The
example below shows this.

1. Currently it is known that 80% of customers shop at store 1 and 20% shop at store 2.
In reviewing a past data suppose we find that out of all customers who shopped at
store 1 in a given week 90% remain loyal for the next week (store one again), 10%
switch to store 2. Out of all customers who shopped at store 2, in a given week 80%
remain loyal for the next week (store 2 again), 20% switch to store 1. What will be
the proportion of customers shopping at store 1 and 2
a) in each of the next two weeks?
b) in the long run?

Lets denote Store 1 by 1 and Store 2 by 2.

V12= (.8 .2) - initial state/ current state probability matrix.

To next weekly shopping period

From one week S1 S2

S1 0.9 0.1

S2 0.2 0.8
 The sum of rows in the transition matrices should be one.
 We have to be consistent in writing the elements.
P11, P22, P33, P44 ---------------------Pnn that represent the primary diagonal show loyalty.
Others switching.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 17

Markov Chain Formula

nth state of a Markov Chain.

Vij (n) = Vij (n-1) x p, or Vij (n) = Vij (0) x (P) n.

Or
Vij (n) = Vij (0) x (P) n.
Where: P = transition matrix
Vij (n) = Vector for period n.
Vij (n-1) = vector for period n-1.

V12 (0) = (.8 .2)

V12 (1) = V12 (0) x P
 .9 .1
 
= (.8 .2) 
 .2


.8


= (.8 x .9) + (.2x.2) (.8x.1) + (.2x.8)

= .72 + .04 .08 +. 16
= 0.76 .24

V12(1) = (.76 .24)

V12(2) = V12(1) x P
= V12(1) x P
= (.76 .24)
 .9 .1
 
= (.8) .2) 
 .2


.8

=(0.732 .268)
b. In the long run (V1 V2) (n) = (V1 V2) (n+1)
n p n+1
 .9 .1
 
(V1 V2) 
 .2


.8

= (V1 V2)
0.9V1 + .2V2 = V1
.1V1 + .8V2 = V2
V1 + V 2 = 1
-.1V1 + .2V2 = 0 
 one is the - ve of the other.
.1V1 + -.2V2 = 0 

.9V1+.2(1-V1) =V1
.9V1 + .2 - .2V1 = V1
.7V1 + .2 = V1
.2 = .3V1
V1 = 2/3
V2 = 1 - V 1

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 18

= 1 - 2/3
V2 = 1/3

In short, the switching over the sum of the switching gives us the long run state.

To

S1 S2

From S1 .9 .1

S2 .2 .8

Switch to state 1 Switch to state 2

V1= Switch to state 1  switch to state 2 V2 = Switch to state 1  switch to state 2
.2 2 .1 1
 
.2  .1 3 .2  .1 3
= =

2 1
(V1 V2) =  
3 3

In the long run 67 of the customer will shop in store 1 and 33% in store 2.
Prediction: Long run - only the transition matrix.
At specified time - the transition matrix and state vector.
Hence, unless the transition matrix is affected, the long run state will not be
affected. Moreover, we cannot know the number of years, weeks, or periods to
attain the long run state, point but we can know the share.

Absorbing Markov Chain

It is a special type of Markov chain in which at least one of the states eventually
doesn’t lose members. We call such a state absorbing because it can absorb
members from other states, but doesn’t give up any of its members.

For example, if we take the above example and change the transition matrix

S1 S2
S1 1 0
S2 .2 .8
The state S1 (store 1) in absorbing

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 19

In short:
Consider a Markov chain with n different states {S1, S2, and S3 --- Sn}.
The ith state Si is called absorbing if Pii = 1. Moreover, the Markov chain is called
absorbing if it has at least one absorbing state, and it is possible for a member of
population to move from any non-absorbing state to an absorbing one in a finite
number of transitions.

Remark: Note that for an absorbing state Si, the entry on the main diagonal p must
be Pii = 1 and all other entries in the ith row must be 0.

To
E.g. a.
S S S
1 2 3

 S 0.4 0 0.6
 1 
from S 2 0 1 0  Absorbing Markov Chain
 
S
 3 0 0. 5 0. 5 
To
E.g. b.
S S S
1 2 3

 S 0.4 0 0.6
 1 
fromS 2 .5 15 0  has no absorbing states.
 
S
 3 0 15 . 5 
To
S S S S
1 2 3 4

S1 .5 15 0 0
 
fr o m  S 2
 0 1 0 0
6
The second state is absorbing.

S 3
0 0 .4

S 4 0 0 5 .5 
However the corresponding Markov chain is not observing. Because there is no way
to move from state 3 or state 4 to state 2.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 20

A Markov chain is absorbing it has at least one absorbing state, and if from every
state it is possible to go to an absorbing state (not necessarily in one step).

Exercises

1. A division of the ministry of public health has conducted a sample survey on

the public attitudes towards the use of condoms. From the results of the survey the
department concluded that currently only 20% of the population uses condoms and
every month 10% of non-users become users, where as 5% of users discontinue
using.
Required

a. Write the current transition matrices.

b. What will be the percentage of users from total population just after two
months?
c. What will be the proportion of the non users and users in the long run?

Solution

Let. U - Stands for users, and N- stands for nonuser

1. Initial state VUN (0) = 0.2 0.8

To the next month

From one month Users (U) Non Users (N)

Users (U) .95 .05

Non Users (N) .10 .90

2. V (1) UN = V (0) UN x P
 .95 .05
= 0.2 0.8  .10 .90
=
(0.27 0.73)

V (2) UN = V (1) UN x P
 .95 .05
= 0.27 0.73  .10 
.90
= (.3295 0.6705)

3. VU VN = (? ?)

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 21

switchtoU switchtoN
Switchtou  SwitchtoN Switchtou  SwitchtoU
VU = VU =
.1 .05
  0.65   0.33
.15 .15

VU VN = 0.67 0.33
VUN (n) = 0.67 0.33

2. A city has two suburbs: suburb x and suburb y. Over the past several years, the city
has experienced a population shift from the city to the suburbs, as shown in the
table below.
To the next year
From City (C) Suburb x (X) Suburb y (Y)
one year City (C) .85 .07 .08
Suburb x (X) .01 .96 .03
Suburb y (Y) .01 .02 .97

In 20xo, the city had a population of 120,000, suburb x had a population of 80,000,
and suburb by had a population of 50,000. Assuming that the population in the
metropolitan area remains constant at 250,000 people,
a. How many people will live in each of the three areas in 20X 2?
b. How many people will live in each of the three areas in the long run?

Solution.

Let C stands for the city

X stands for the suburb X.
Y stands for the Suburb y. C= 120,000 - 120,000/250,000 = 0.48
x = 80,000 - 80,000/250,000 = 0.32
y = 50, 000 - 50,000/250,000 = 0.20
250,000 1.00

Initial state V(0)cxy (0.48 0.32 0.20)

The transition matrix. From one year

 C X Y 
 
C .85 .07 .08 
P= X .01 .96 .03 
 
Y .01 .02 .97 


Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 22

 
 
 .85 .07 .08
V(1)cxy = V(0)cxy x p (.48 .32 .20)  .01 .96 .03
 
 .01 .02 .97

V (1)cxy = (.4132 .3448 .2420)

 
 
 .85 .07 .08
V(2)cxy = (.4132 .3448 .2420)  .01 .96 .03
 
 .01 .02 .97
V(2)cxy = (.3571 .3648 .2781)

Thus, in 20X2, 89,275, 91,200 and 69,525 people will live in the city, suburb x and
suburb y respectively.

longrun
b. n p n+1
 .85 .07 .08
 
(Vc Vx Vy)  .01 .96 .03 (Vc Vx Vy)
 
 .01 .02 .97

.85C + .01x + .01y = C

.07C + .096x + .02y = x
.08C + .03X + .97y=y
c + x + y =1 Vc Vx Vy = 1
-.15C + .01x + .01y = 0
.07c - .0yx + .02y = 0
.08C + .03x - .03Y = 0
X = 1-C-Y

.07 - .04 (1-c-y) + .024 = 0

.07 - .04 + 0yc + 04y) + .02y = 0
(.07c + .0yc) - 04 + (04y + .02y) = 0
.11c+.06y - .0y = 0 --- .08c - .03 (1-c-y) + .03y = 0
.08c - .03 + 03c + 03y + .03y = 0
(.08c - .03c) + 03 + (03y + .03y) = 0
.05C +.03 - .06y = 0 --- (2)
.11c + .06y - .4 = 0

.05C - .06Y + .03 = 0

.16C - .01 = 0
.16C = .01

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 23

.01
C=
.16
C = 0.0625 .11 (.0625) + .06y - .06y - .04 = 0
.006875 + .06y - .04 = 0
.06y = .033125
y = 0.5521

C+X+y = 1
.0625+x+.5521=1
0.6146+x=1
X = .3854

(Vc Vx Vy) = (.0625 .3854 .5521)

In the long run 15,625, 96,350 and 138,025 people will live in the city suburban X
and suburban respectively.

3. A population of 100,000 consumers makes the following purchases during a

particular week: 20,000 consumers make the following purchases during a
particular week: 20,000 purchase Brand A, 35,000 Brand purchase B and 45,000
purchase neither brand. From a market study, it in estimated that of those who
purchase Brand A, 80% will purchase it again next week, 15% will purchase Brand
B next week, and 5% will purchase neither brand. Of those who purchase B, 85%
will purchase it again next week, 12% will purchase brand A next week, and 3%
will purchase neither band. Of those who purchased neither brand, 20% will
purchase as A next week, 15% will purchase Brand B next week, and 65% will
purchase neither band next week. If this purchasing pattern continues, will the
market stabilize? What will the stable distribution be? Yes Y A Vb Vc = (.4 .5 .1)

4. In a certain college class, 70% of the students who receive an “A” on the current
examination will receive an “A” on the next examination. Moreover, 10% of the
students who do not receive an “A” on the current examination will receive an “A”
on the next examination. Assuming that this pattern continues, what is the stable
matrix?
VA VA1 = (.25 .75)

5. A vigorous television advertising campaign is conducted during the football season

to promote a well-known brand X shaving cream. For each of several weeks, a
survey is made and it is found that each week 100% of those using brand X continue
to use it. It is also found that of those not using brand x, 20% switch to brand X
while the other 80% continue using another band.

Mathematics for Management (MGMT 121)

 Chapter Two: Matrix Algebra and its Applications 24

a. Write the transition matrix, assuming the transition percentages continue hold for
succeeding weeks.
b. If 20% of the people are using brand X at the start of the advertising campaign,
what percentage will be using brand X one week later? Two weeks later?
c. What portion of the market will be using brand X area the end of the season,
assuming the transition matrix remains the same? Find the Steady-state matrix)

soln.
X x
A. X 1 0
X .2 .8

1 0
B . V(n)xx1 = V(0)xx1p = (.2 .8) 
 .2 
 .8 

V(n)xx1 = (.36 .64)

1 0
V(2)xx1 = V(1)xx1xp = (.36 .64) 
 .2 
 .8 

= (.488 .512)

1
C. Vx = =1
1 0
0
Vx' = =0
1 0

Mathematics for Management (MGMT 121)