CHAPTER TWO

MATRIX THEORY AND ITS APPLICATION

2.1. Review on Matrix Algebra
Matrix could be defined in different ways having the central or core idea unchanged. These
differences simply mount to wording. Some of the definitions are:
A matrix is a rectangular array of numbers. Alternatively, matrix is a rectangular array of
numbers arranged in rows and columns. In other words, it is a collection of numbers which are
arranged in rows and columns. Matrix is a common device for summarizing and displaying
numbers or data.
The word “matrices” is the plural form of matrix. Numbers which form matrices are enclosed by
using symbols like parentheses,  or brackets . Higher case letters are used to represent
matrix and small letters are used to represent numbers which form a matrix. Each member in a
matrix is called an element.
Example 2.1
1 0 1
A = 2 B= 2 4
elements
In general, matrix A containing elements aij has the general form:

A= a11 a12 a13 … a1n

a21 a22 a23 … a2n
.
am1 am2 am3 … amn mxn, where

aij is element of a matrix

in aij, i represents the row
j represents the column
mxn represents dimension of matrix, where
m is number of rows and
n is number of columns.

1
In the above matrix, A, a12 is read as " a sub one two" and a12 represents the element (entry) of
the matrix and subscript "12" represents the position of the entry. The entry a 12 is located at the
intersection of row one and column 2. Row represents the horizontal array of numbers and
column represents the vertical array of numbers.

Dimension of a matrix represents the order, shape or size of a matrix. Since matrix "B" above has
two rows and two columns, its dimension is 2x2 (read as two by two) and is called a two by two
matrix.
Example 2.2
These illustrative examples will help you to clarify any ambiguity.
1. [0] is a 1x1 matrix, because it has only one row and one column.
2. 1 2 3 is a 1x3 matrix.
3. 1
2 is a 2x1 matrix.

4. 0 2 4
2 1 0 is a 2x3 matrix.

5. 1 2 3
4 5 6
7 8 9
0 2 4 is a 4x3 matrix.

1: To avoid having to specify which of a pair of subscripts represent the row and, which is the
column, we shall always use ‘the row, column’ order convention, where the 1st subscript “1” in
a12 for example refers to row and the second subscript “2” refers to column.
2: A comma is required to avoid ambiguity if the number of rows or columns exceeds nine. For
example, an entry in the 12th row and 3rd column would be written as a12, 3.

2
Example 2.3
75 82 86
A= 65 70 68
73 76 74
59 80 99
In the above matrix A,
i) The dimension is 4x3.
ii) a12 represents an entry 82(at the intersection of row1 and column 2).
iii) a23 = 68
iv) a35 = no element
Determine the following from the above matrix A.
i) a22 =______________
ii) a31 = _________
iii) a43 =______________
iv) a55= ______________

Application
4. Suppose that a company has six gasoline stations, three in region # 1 and three in region # 2.
January sales volume, in thousands of gallons, is shown for each station as:
Sales in thousands of gallons, January
Station Region #1 Region #2
#1 10 15
#2 12 18
#3 8 12
Change this problem into a matrix form
i) Keeping in mind that rows are stations and columns are regions
ii) Determine a region by station sales matrix

3
2.2. Types of Matrices
a) Vector matrix- This is a matrix which has precisely one row or precisely one column.
From this definition, we can identify two classifications
i) Row matrix is a matrix having only one row. It is a 1xn matrix, and is called a row vector of
dimension n, i.e., 1xn matrix.
Example 2.4
A= [2 1 4] is a 1x3 matrix.
B= [1 2] is a 1x2 matrix.
ii) Column matrix is a matrix having only one column. It is an m by 1 matrix, and is
also called a column vector of dimension m, i.e., mx1 matrix.
Example 2.5
1 4
C= 2 D= 9 2x1 matrix
3 3x1 matrix
Entries in row and column vectors often are referred to as components of the vectors.
b) Square matrix is a matrix having the same number of rows and columns, i.e., for two or more
matrices to be square matrices, they should have to have the same shape or size. If the dimension
of a matrix is mxn, for the square matrix, m = n, mxn =mxm = nxn.
Example 2.6
1 0 2
C= 3 1 2
4 3 7 is a 3x3 matrix.

D= 1 0
4 3 is a 2x2 matrix.
c) Equal matrices – For two or more matrices to be said equal matrices, necessarily they should
have the same shape and the same corresponding entries.

2 3 5 2 6/2 6-1
G= 5 0 6 H= -1+6 0 6
1 2 1 is a 3x3 matrix. 1 4/2 2-1

4
Matrices G and H are equal matrices, because the two are 2x2 matrices and their corresponding
entries are the same.
d) Identity matrix or Unit matrix is a square matrix where elements along the primary or main
diagonal all equal to one (1), while all other elements are equal to zero. Main diagonal is a line
that runs from upper left to lower right. Identity matrix is denoted by I.
Example 2.7
Main diagonal
1 0 0 1 0
A= 0 1 0 3x3 matrix B= 0 1 2x2 matrix
0 0 1

(e) Zero matrix is a matrix whose elements are all zeros and is designated by O.
Example 2.8
V= 0 0 0 W= 0 0 0
0 0 0 0 0 0
The above matrices are zero matrices.
(f) Scalar Matrix is a matrix with the same constant number on its main diagonal and
zeros in all other locations.
Example 2.9

A= 5 0 0 B= ½ 0 0
0 5 0 0 ½ 0
0 0 5 0 0 ½
The above matrices are scalar matrices.
(g) Diagonal matrix is a matrix with value other than zero in its main diagonal and all other
elements equal to zero.
Example 2.10

C= ½ 0 0
0 -2 0
0 0 3

5
From the above discussion, we can infer that:
 Identity matrix can be both scalar and diagonal matrices
 Equal matrices are square matrices, but the vice versa may not necessarily be
true.
 Scalar matrix can be diagonal, but the vice versa may not necessarily be true.

2.3. Matrix operations and their properties on matrices

2.3.1. Addition and Subtraction on Matrices
Matrices are added or subtracted by adding or subtracting the corresponding entries. Since
entries must correspond, matrices must involve the same dimensions. Therefore, the sum or
difference of two or more matrices of the same size can be obtained by adding or subtracting the
corresponding elements of two or more matrices. This is possible only if the matrices have the
same size.
Example 2.11
Let A= 2 3 B= 2 3 1 C= 2 -1
4 -1 0 1 4 0 1
Find:
a) A + B
b) A + C
What is the sum and difference of the following two matrices?
H= ½ 6
4 5/2 J= 4 3
0 1 2/5 6

Properties of Addition and Subtraction on Matrices

Because we add two matrices by adding their corresponding elements, it follows from the
properties of real numbers that matrices of the same dimension are commutative and associative
relative to addition.
i) Commutative property

For two matrices with the same size, if interchanging of the positions of matrices has no change
on the sum, then addition is communicative. Thus, for two matrices A and B, A+B= B+A.

6
Example 2.12
A= 2 4 B= -2 6
3 5 1 2
Prove that A+B =B+A
ii) Associative Property
For three matrices, A, B and C with the same dimension, (A+B) + C =A+ (B+C) is called
associative property.
iii) Identity Property
For any matrix A, there is a corresponding matrix –A, such that
A +(-A) = 0 = (-A) +A. The negative of a matrix A, denoted by –A, is a matrix
with elements that are the negative of elements in A. Thus, if

A= a b then, -A = -a -b
c d -c -d

2.3.2. Multiplication of Matrices

Matrix multiplication is a specialized form of multiplication devised for very particular reasons.
a) Scalar Multiplication
In matrix algebra, an ordinary number is called scalar. The product between a real number
(scalar) and a matrix is called scalar product to multiply a matrix by a scalar, each entry of a
matrix is multiplied by a scalar. Therefore, the product of a number K and a matrix M, denoted
by KM, is a matrix formed by multiplying each element of M by K.
Example 2.13

B= 2 1 2 = 2 4 C=-1/2 3 -1 0 = -3/2 1/2 0

-3 4 -6 8 -2 0 3 1 0 -3/2

7
Example 2.14
Application: W/t Y and Ato X are sales people in a new car agency that sells only two models.
August was the last month for this year’s models, and next year’s models were introduced in
September. Gross birr sales for each month are given in the following matrices.
August sales (in 000) September sales (in 000)
Compact Luxury Compact Luxury
Y Br 54 88 =A 228 368 =B
X 126 0 304 322

i) What were the combined birr sales in August and September for each person and each model?
ii) What is the average birr sale in August and September for each person and each model?
iii) If both sales people receive 5% commission on gross birr sales, compute the commission for
each person for each model sold in September?
iv) What was the increase in birr sales from August to September?
b) Inner Product
Inner product, which is also called dot product, refers to the multiplication of elements in row
vector with column vector. The result of inner product is always a scalar. Inner product is
defined when the number of row vector elements is equal to number of elements of column
vector.

Example 2.15
Let A = 1 7 5 a row matrix

B= 2
4 a column vector
1
What is the inner product of A and B? 8
Exercise: What is the product of C= 2 6 3 and D= 1
4

8
Example 2.16
Application: A dealer of color television sets has five 26 inch sets, eight 20 inch sets, four 18
inch sets, and ten 12 inch sets. The 26 inch sets is sold for $650 each, the 20 inch sets is sold for
$550 each, the 18 inch sets is sold for $500 each and the 12 inch sets is sold for $300 each.
Express the total selling price of his television stock as the product of two matrices.

c) Matrix Product
The product of two matrices A and B, AB is defined if and only if the number of rows of matrix
B is the same as number of columns of A. If this is satisfied, the two matrices are said to be
conformable for multiplication. The dimension of the resulting matrix, AB is the same as number
of rows of matrix A and number of columns of matrix B.
Therefore, given matrices A and B having dimension of
A (mA x nA ) and B (mB x nB ),

A.B= C (new matrix).

(mA x nA )(mB x nB ) = C (mA x nB )

If nA= mB, the matrices are conformable for multiplication.

Generally, if A is axb matrix, and B is cxd matrix, then AB (the new matrix) is axd matrix.
Example 2.17

For A= 2 8 1 and B= 1 7
3 6 4 9 -2
AB =? BA=? 6 3

-3
Example 2.18. A= 4 B= 8 2 then, AB=?
7
8 5 2
Example 2.19. E= 6 1 F= 8 then, EF =? FE=?
9 4 5

9
Exercise. G= 8 5 3
6 1 H= 7 then, GH =? HG=?
9 4

Example 2.20
Application. A personal computer retail company sells five different computer models through
three stores located in a large metropolitan area. The inventory of each model on hand each in
each model is summarized in matrix M. Whole Sale (W) and retail (R) values of each model
computer are summarized in matrix N.

Model
A B C D E W R
4 2 3 7 1 Store 1 Birr 700 Birr480 A
M= 2 3 5 0 6 Store 2 N= 1,400 1800 B
10 4 3 4 3 Store 3 1,800 2,400 C
2,700 3,300 D
3,500 4,900 E
i) What is the retail value of the inventory at store 2?
ii) What is the whole sale value of the inventory at store 3?
iii) Compute MN
Example 2.21: A manufacturing firm which manufactures office furniture finds that it has the
following variable costs per unit in dollar/unit.
Desks Chairs Tables Cabinets
Material 50 20 15 25
Labor 30 15 12 15
Overhead 30 15 8 20

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

10
Properties of Multiplication
 Scalar multiplication is distributive and associative with respect to addition.
K (A+B) = KA+KB, where K is scalar.
 Scalar multiplication also holds true for associative property.
K. (A.B) = (K.A).B
 Matrix multiplication is not commutative, i.e., A.B ≠B.A. But, under exceptional
conditions it may be commutative:
i) When the two matrices are equal, i.e., A=B, hence A.B=B.A.
ii) When a matrix is multiplied by identity matrix, AI=IA.
 Matrix multiplication is associative. i.e., A.(B.C)= (A.B).C. But the three matrices should
be conformable.
 Matrix multiplication by an identity matrix results in original matrix.
2.4. Solving System of Linear Equation Using Matrix and Matrix Application
In solving systems of equations, we are interested in identifying values of the variables that
satisfy all equations in the system simultaneously. The solution sets for a system of linear
equations may be a null set, a finite set or an infinite set.
There are quite a few solution procedures which may be used in solving systems of linear
equations. These are Gaussian elimination method, inverse method and determinant method. In
this section we will discuss the first two.
An mxn system of equations having the form:
a11x1 + a12 x2 + a13x3 . . . a1n xn = b1
a21 x1 + a22 x2 + a23 x3 . . . a2n xn = b2
. . . . .

am1x1 + am2 x2 + am3 x3 . . . amn xn = bm , can be written in a matrix form

as:
a11 a12 a13 . . . a1n x1 b1
a21 a22 a23 . . a2n x2 b2
. . . . . .
. . . . . = .
. . . . . .
am1 am2 am3. . . amn xn bm
Coefficient of equations Variables constant terms
In general, a linear equation of the form a 1x1 + a2 x2 + a3x3 + . . . a n xn = b can be written in a
matrix form as:
11
a1 a2 a3 . . . an x1 = b, this form is called compact form. Hence, ax = b.
x2
.
xn

Example 2.22
2x1 + 5x2 = 9
3x1 + x2 = 3
-2x1+ x2= 1 , then the matrix form of this system of linear equation is:

2 5 x1 9
3 1 x2 = 3
-2 1 1

Example 2.23: Express the following system of equations in matrix form:

2x+3y+ 4z =7
4y = 2+5z
3z-2x+6= 0
2.4.1. Solving System of Linear Equations Using Gaussian Elimination Method
The procedure begins with the original system of equations and transforms it, using row
operations, into an equivalent system from which the solution may be read directly.
These row operations are:
1. Interchanging two rows
2. Multiplying or dividing a row by a non zero constant
3. Adding (or subtracting) a constant multiple of one row to (or from) another row.
Steps of Gaussian Method
a. Write the system of linear equations in augmented form (i.e in the form of A/B where A
is coefficient matrix and B is constant matrix).
b. Apply elementary row operations to transform the original augmented matrix into another
one ( i.e A/B into I/C)
c. Decide the values of variable matrix, and the values are determined when A is reduced to
identity I matrix. Then the final augmented matrix is called reduced matrix and C is
solution value matrix.

12
 Some guidelines about row reduction procedure
1. Use row operations to obtain a top entry in the first column equal to 1
2. Add or subtract the appropriate multiples of the first row to the other rows so that the
remaining entries in the first column become zero.
3. Without distributing the first column, use row operations to make the second entry in the
second column equal to 1. Then add or subtract suitable multiples of the second row to
the other rows to obtain zeros in the rest of the second column.
4. Without disturbing the first two columns, use row operations to make the third element in
the third column equal to 1. Then use the third row to obtain zeros in the rest of the third
column.
5. Continue the process column by column until the transformed or reduced form (I/B) of
the matrix is obtained, with the identity matrix I on the left of the vertical line.

Example 2.24: Use the method of row reduction to solve the following system of linear
equations:
2x-3y+4z=13
x+y+2z= 4
3x+5y-z= -4
Example 2.25: A furniture manufacturer makes chairs, coffee tables, and dining room tables.
Each chair requires 2 minutes of sanding, 2 minutes of staining and 4 minutes of varnishing.
Each coffee table requires 5 minutes of sanding, 4 minutes of staining and 3 minutes of
varnishing. Each dining room table requires 5 minutes of sanding, 4 minutes of staining and 6
minutes of varnishing. The sanding and varnishing benches are each available for 6 hours per
day and the staining bench is available for 5 hours per day.
a. Develop the system of linear equations for the situation
b. Write the system of linear equation in matrix multiplication form
c. How many of each type of furniture can be made daily?

2.4.2. Solving System of Linear Equations Using the Inverse of a Matrix

13
For some matrices there can be identified other matrix called the multiplicative inverse matrix, or
more simply the inverse. The relationship between a matrix A and its inverse (denoted by A -1) is
that the product of A and A-1 in either order, results in the identity matrix or AA-1=A-1A=I
Note: I. For a matrix A to have an inverse it must be square
II. The inverse of A will also be square and of the same dimension as A.
III. Not every square matrix has an inverse.
Before using inverse of a matrix to find a solution of system of equations, let us see how to
determine the inverse of a matrix. There are several methods for determining the inverse of a
matrix. However, in this material, we are limited to the Gaussian method of determining the
inverse.
To determine the inverse of an (m x m) matrix A:
a) Augment the matrix A with an (m x m) identify matrix (i.e. , A/I).
b) Perform row operation on the entire augmented matrix so as to transform A into an (m x
m) identify matrix. The resulting matrix will have the form (I/A-1); where A-1 can be read
from right of the vertical line.
Example 2.26: Find the inverse of the matrix:

A= 3 7
2 5

Example 2.27: Find the inverse of the (3 x 3) matrix B

1 2 0
B= 1 0 -1
-1 3 2

Note: If a matrix A does not have an inverse it will not be possible to transform A into an
identity matrix.

Example 2.28: Consider the following matrix and determine its inverse.

14
C= 2 -4 6
6 1 5
1 -2 3

Now let us see how the matrix inverse can be used to determine the solution set for a system of
equations. Given a system of equation of the form AX=B, where A is a square matrix of
coefficients, both sides of the matrix equation may be multiplied by A-1 yielding:
A-1A X=A-1B
Because, A-1.A=I, rewrite the above equation as
IX= A-1.B and since I times any matrix is the matrix itself, the following results:

X=A-1.B

This result indicates that the solution vector for the system of equations can be found by
multiplying the inverse of the matrix of coefficients (coefficient matrix) A by the vector of right
side constants (constant matrix) B.
Example 2.29: Consider the system of equations:
3x1+7x2=27
2x1+5x2=19
This system of equations may be written in matrix form as AX=B

3 7 x1 = 27
2 5 x2 19

To solve this system of equations by the inverse method, we must determine A -1 conveniently; A
is the matrix examined in the example 2.26 above. A -1 was computed in that example and the
inverse was A-1 = 5 -7
-2 3 .
Therefore, the solution vector X is calculated as:

15
X= A-1.B

X= 5 -7 27 = 2
-2 3 19 3

X1 = 2
X2 3

The solution to the system of equations is X1=2 and X2=3

Example 2.30 : Solve the system of equations below using inverse of the matrix
x1+2x2 =1
x1-x3 =2
-x1+3x2+2x3= -5

Exercises: (Use matrix concepts to solve the problems)

1. Solve the following system of linear equations.
x +2y + 3z =3
2x +5y +7z =6
3x +7y+8z=5
2. Melba Carpet Co. has an inventory of 1500 square yards of wool and 1800 square yards of
nylon to manufacture carpet. Two grades of carpeting are produced. Each roll of superior grade
carpeting requires 20 square yards of wool and 40 square yards of nylon. Each roll of quality
grade carpeting requires 30 square yards of wool and 30 square yards of nylon. If Melba would
like to use all the material in inventory, how many rolls of superior and how many rolls of
quality carpeting should be manufactured?

3. Jonfa invested a total of Br. 10,000 in three different saving accounts. The accounts paid
simple interest at an annual rate of 8%, 9% and 7.5% respectively. Total interest earned for the

16
year was Br. 845. The amount in the 9% account was twice the amount invested in the 7.5%
account. How much did Jonfa invest in each account?
4. Attendance records indicate that 80,000 South Koreans attended the 2002 world cup at its
opening ceremony. Total ticket receipts were $3,500,000. Admission prices were $37.50 for the
second class and $62.50 for the first class. Determine the number of South Koreans who attended
the football game at first class and second class.

2.5. Markov Chains

If a system evolves from one state to another in such a way that we are able to assign
probabilities other than 0 or 1 to the occurrences of future states, then the progression of the
system through a sequence of states is called a stochastic process. One of the special kinds of
stochastic (probabilistic) process is called Markov Chain

Markov Chain is defined as a sequence of experiment, trials or observations such that the
probabilities for the next state are completely determined by the present state. Alternatively,
Markov chain is a model which is useful in studying the evolution of certain system over
repeated trials or experiments.
Characteristics of Markov Chains
1. The situation occurs in a chance environment consisting of two or more possible
outcomes that occur at the end of a well defined fixed period.
2. Since the situation involves a multi period case, the probabilities of the evolution tend to
be transitional probabilities. Transitional probability (matrix) is a square matrix having
no negative entries and the sum of entries in each row equals to 1.
3. The process is observed after each transition and is governed by transition probabilities.
4. If a transition probabilities of a system from one period to another remains constant and
the outcomes of the system in a particular time period depends only on the previous time,
the Markov chain process is called Homogeneous Markov Chain.

Important Assumptions
 The system has finite number of states

17
  The system condition or state in any given period depends only on its state in the
preceding period and transition probabilities.
 Transition matrix (probability) is constant over time and the process of keeping these
transition probabilities is called stationary Markov Chain.
 Changes in the system may occur once and only once each period.
Example 2.31: Currently it is known that 80% customers shop at store 1 and 20% shop at store
2. In reviewing 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. On the
other hand, out of all customers who shopped at store 2, in a given week 80% remains 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 in each of the next two weeks? What will be the long run market share?
Example 2.32: A population of 100,000 customers make the following purchases during a
particular week: 20,000 purchase Brand A, 35,000 brand B and 45,000 purchase neither brand.
From a market study, it is 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 brand. Of those who purchase neither brand, 20% will
purchase A next week, 15% will purchase brand B next week and 65% will purchase neither
brand next week. If this purchasing pattern continuous, will the market stabilize? What will the
stable distribution be?
Example 2.33: Find the steady state for the transition matrix 0.5 0.5
1 0

18