Chapter Two

Matrix Algebra and its Application in Business

2.1 Introduction to Matrices
Whenever you deal with a set of information about one or more similar situations, there should
be concern for organizing them in such a way that they are meaningful and can be readily
identified. Summarizing information in a tabular form can serve this purpose. Hence you can say
that matrix is a common device for summarizing and displaying information or numbers.
Therefore, you have the following definition.

Definition 2.1 A matrix is a rectangular array of real numbers arranged in m-rows and n-
columns.

It is symbolized by a bold face capital letter with its elements enclosed within a parentheses or
bracket as shown below.
 a11 a12  a1n  a11 a12  a1n 
  a
 a 21 a 22  a 2 n   a 22  a 2 n 
A or A  21
in which aij are real numbers
       
   
a  
 m1 a m 2 a mn  a m1 a m 2  a mn 
Each number appearing in the array is said to be an element or component of the matrix.
Element of a matrix are designated using a lower case form of the same letter used to symbolize
the matrix itself. These letters are subscripted 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 raw and jth column of the arrangements of the matrix A .
The number of rows (m) and the number of columns (n) of the array give its order or its
dimension and denoted by “m x n” and read as “m by n”.
The following are examples of matrices
1 7 
 
a) A   5 3  This is 3 x 2 matrix where element a12  7, a21  5, a32  2, etc.
 4 
 2  32
Row poison
Column poison
Indicator
indicator x 44  45
 1 5 9 15 
  x  30
b) X   2 6 10 20  This is 4  4matrix where some elements are: 34
 3 7 11 30  x 42  8
 
 4 8 12 45  44 x32  7

Row poison Column poison
Indicator indicator

1
2.2 Types of Matrices
There are deferent types of matrices. These are
1.Vector matrices – are matrices which consist of just one row or just one column. The
dimension of the former is m x 1 matrix and the latter is 1 x n matrix.
1.1 Row vector or matrix is a row matrix with dimension 1 x n.

For example W  (  1 0 6 )1 x 3 is a row matrix.

1.2. Column vectoror matrix is a column a matrix with dimension m x 1.
 0
 
For example Z   20  is a column matrix.
3 
  31
2.Square matrix is a matrix that has equal number of rows and columns. It is also called nth
order matrix or a square matrix of order n. If a matrix is not square then it is
called rectangular matrix.
1 2
For example X    is a square matrix of order 2 and
 3 4  22
 2 5 7 
 
Y   3  9 12  is a square matrix of order 3.
 4 8  33
 0
3. Null or zero-matrix is a matrix that has zero for every entry. It is generally denoted by capital
letter O.
 0 0
 
For example O   0 0  is a zero matrix of dimension 3  2 but
 0 0
  32
 0 0
O    is a zero matrix of order 2.
 0 0  22
For square matrix, the diagonal from the top left to the bottom right corner is called the principal
diagonal or main diagonal or primary diagonal.
 a11 a12  a1n 
 
 a 21 a 22  a 2 n 
A is square matrix of order n and the principal diagonal is the diagonal
   
 
a 
 n1 a n 2  a nn 
that contain the elements a11 , a22 , a33 ,, ann .
4. Identity or unit matrix a square matrix in which all of the primary diagonal entries are ones
and all of the off diagonal entries are zeros. Identity matrix of order n is
denoted by In.

2
 1 0 0 0
1 0 0  
1 0   0 1 0 0 
For example I 2    , I3   0 1 0  , I4   are identity matrices of
 0 1  0 0 1 0 0 1 0
   
 0 1 
 0 0
order 2, 3, and 4, respectively
5. Scalar matrix is a square matrix where elements on the primary diagonal are the same.
 2 1 0 
 
For example  0 2 1  is a scalar matrix.
 2 1 2 
  33
“An identity matrix is a scalar matrix but a scalar matrix may not be an identity matrix”.

2.3. Matrix operations and Properties

Similar to arithmetic operations, in this section you will deal with matrix operation which
involves addition, subtraction, scalar multiplication, and product of two matrices.
Addition of matrices
Two matrices of the same dimension are said to be conformable for addition.
Definition 2.2 If A and B are two matrices, each of dimension m x n, then the sum of A and B is
the m x n matrix whose elements are obtained by adding the corresponding
elements:
 a11 a12  a1n   b11 b12  b1n   a11  b11 a12  b12  a1n  b1n 
     
 a 21 a 22  a 2 n   b21 b22  b2 n   a 21  b21 a 22  b22  a 2 n  b2 n 
A B    
             
     
 a   b   a  b    
 m1 a m2 a mn   m1 m 2
b b mn   m1 m1 a m2 bm2 a mn bmn 

1 3   7 9 
Example 2.1 a) If A   , B    , then
 2 4  8  10 
1 3   7 9   1 7 39  8 12 
A  B            
 2 4   8  10   2  8 4  (10)   10 6 
 2 7 2 8 7 
b) C   , D   , then
 4 6 9  6 4
 2 7 2  8 7 
C  D        is not defined.
 4 6 9  6 4
Since the two matrices C and D aren’t conformable for addition because they aren’t of the same
dimension since the former is 2  3 and the later is 2 2 matrix.

Properties of matrix addition

The operation of adding two matrices that are conformable for addition has these two basic
properties.

3
 1. A + B = B + A (Commutative law of matrix addition)
2. (A + B) + C = A + (B + C) (Associative law of matrix addition)

Multiplying a matrix by a constant (scalar multiplication)

A matrix can be multiplied by a constant. If you multiply each component of the matrix with
constant you have a new matrix of the same dimension as the original matrix. This is called
scalar multiplication of matrices.
Definition 2.3 If k is any real number and A is an m x n matrix, then the product kA is defined
to be the matrix whose components are given by k times the corresponding
component of A . That is
 a11 a12  a1n   ka11 ka12  ka1n 
   
 a 21 a 22  a 2 n   ka21 ka22  ka2 n 
kA  k  
        
   
 a   
 m1 a m 2  a mn   kam1 kam 2  kamn  mn
Example2.2 a) If A  (6 5 7) and k  2 , then 2 A is
2 A  2(6 5 7)  (2  6 2  5 2  7)  (12 10 14)
 1 2 4 
  2 2
b) If B    3 2 3  and k   , then  B is
 0  3 3
 5 5 
 1 2 4   2 3 4 3 8 3 
2 2   
 B    3 2 3  2 4 3 2 
3 3
 0 5  5   0  10 3 10 3 
Properties 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 m x n 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
Subtraction of matrices
You can also find the difference between two matrices of the same dimension as given in the
following definition.

Definition 2.4 If A and B are two matrices, of dimension m x n, then the difference of A and B
is the m x n matrix whose elements are obtained by subtracting the corresponding
elements:

4
  a11 a12  a1n   b11 b12  b1n   a11  b11 a12  b12  a1n  b1n 
     
 a 21 a 22  a 2 n   b21 b22  b2 n   a 21  b21 a 22  b22  a 2 n  b2 n 
A B    
             
     
 a     
 m1 a m 2  a mn   bm1 bm 2 bmn   a m1  bm1 a m 2  bm 2  a mn  bmn 

The subtraction of two matrices can be defined in terms of addition of matrices and scalar
multiplication. If two matrices A and B each of dimension m x n, then the difference of A and
B is the m x n matrix such that
A  B  A  (1) B .
1 2   0 1 
Example2.3If A    and B    , then find A  B and B  A ?
3 4  2 5
 1 2   0 1   1 0 2 1   1 1
A  B             , whereas
 3 4   2 5   3  2 4  5   1 1 
 0 1   1 2   0 1 1 2   1 1 
B  A              ( A  B )
 2 5   3 4   2  3 5  4   1 1 
Note that: A – B B – A
Exercise: Perform matrix addition, and scalar multiplication
In Each question compute the given operations
 4 1 3   8  12 22 
1) Let D    and E    , then find
 6 7 9   7  9 15 
2
a) D  E b) D
3
 2  1   1  5 
2) Let A    and B    , then find
3 4  7 2 
1 4
a) 2 A  3B b) B  A
5 3

Matrix by matrix multiplication

Before you define the product of two matrices you need to observe carefully that
i) If A and B are two matrices, the product AB is defined if and only 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 p. If this requirement is met, A is said to be conformable to B for
multiplication.
ii) If i) is satisfied the matrix resulting from the multiplication has dimension equivalent to the
number of rows in A and the number columns in B . Hence if A is of dimension n x m and
if B is of dimension m x p, then the product C  AB is of dimension n x p.
A of dimensions B of dimension

5
n x m m x p

Must be the same

Dimension of C  AB
nxp

iii) If i) and ii) are satisfied in multiplying two matrices A and B , to obtain the element of the
resulting matrix in the ith row and jth column of C  AB , we take the sum of the product of the
corresponding elements of the ith row of matrix A and jthcolumn of matrix B as shown below:
In order to define a product matrix i.e. matrix C the corresponding elements of A and B as per
the requirement will be defined, and the element for matrix C will be defined in the following
manner:
 a11 a12  a1n   c11 c12  c1 j  c1 p 
   
 21
a a  a 
2n  
b b  b  b   c c  c  c 
22 11 12 1j 1p
  21 22 2j 2p

      b21 b22 b2 j b2 p      
AB    
  
 ai1 ai 2  ain        c c  c  c
   i1 i2 ij ip

      bn1 bn 2 bnj bnp       
 a   
 m1 a m 2  a mn   c m1 c m 2  c mj  c mp 
Where
cij  ai1b1 j  ai 2 b2 j    aik bkj with i  1, 2, 3,, m , k  1, 2, 3,, n and j  1, 2, 3,, p .
You can observe that the element in the first row and first column c11 is obtained by adding the
product of the corresponding elements of the first row of A and the first column of B resulting:
c11  a11b11  a12b21    a1n bn1
 1 7 
 2 3 4  
Example 2.4 If A    and B   0 8  , then find AB and BA ?
 6 9 7  23  5 1 
  32
Solution Now to find AB , the product is defined as the number of columns of A and number of
rows of B are equal to 3. Furthermore, you can observe that the resulting matrix is 2  2 matrix,
which is the number of rows of A and the number of columns of B . Therefore, the product is
performed as follows.
 1 7 
 2 3 4    2(1)  3(0)  4(5) 2(7)  3(8)  4(1)   18 42 
AB    0 8       
 6 9 7  5 1   6(1)  9(0)  7(5) 6(7)  9(8)  7(1)   29 121 
 
Similarly, BA is defined and the resulting matrix has dimension 3  3.
 1 7 
  2 3 4 
AB   0 8  
 5 1  6 9 7 
 

6
   1(2)  7(6)  1(3)  7(9)  1(4)  7(7)   40 60 45 
   
  0(2)  8(6) 0(3)  8(9) 0(4)  8(7)    48 72 56 
 5(2)  1(6) 5(3)  1(9) 5(4)  1(7)   16 24 27 

Note that from the above example AB  BA

Some Properties of Matrix Application

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 and addition, then
i) A( BC )  ( AB)C (Associative properties of matrix multiplication)
ii) A( B  C )  AB  AC (Left distributive property)
i) ( A  B)C  AC  BC (Right distributive property)

2. On the other hand, the commutative law of multiplication doesn’t apply to matrix
multiplication. That is, for two matrices A and B , it is not generally true that AB equals BA .

3. In many instances for two matrices, A and B , the product AB may be defined while the
product BA is not and vice versa.
In some cases, AB  BA . In such cases A and B are said to be commute. For example if
1 1  2 2
A    and B    , then
1 1  2 2
1 1 2 2   4 4   2 2 1 1  4 4 
AB        and BA       
1 1 2 2   4 4   2 2 1 1  4 4 
1. Another unusual property of matrix multiplication is that the product of two matrices can be
zero even though neither of the two matrices themselves is zero. For example if
 3 0 0 0 0 0
   
A   3 0 0  and B   7  10 4  , then
1 0 0  8 3 2 
  
 3 0 0  0 0 0  0 0 0
    
AB   3 0 0  7  10 4    0 0 0  .
1 0 0  8 3 2   0 0 0 
 
You can’t conclude from the result AB  0 that at least one of the matrices A or B is a zero
matrix.
2. Also you cannot in matrix multiplication, necessarily conclude from the result AB = AC
imply that B= C even if A  0. Thus the cancellation law doesn’t hold, in general, in matrix
multiplication.
 1 3  4 3  1 0
For example if A    , B    and C    , then
 2  6  2 5 3 6 
 10 18 
AB  AC    . However, B  C .
  20  36 

7
Exercise: Computation of matrix multiplication
In computation of matrix multiplication take care on the dimensions
 1 1 
 2  1  
1) Let A    and B   2 4  , then find
 3 2  0 1
 
a) BA b) AB , whenever possible.

2.4. Inverse of a Matrix

As in the case of the real number system, you are not to have the multiplicative inverse of any
matrix directly. However, you can have the inverse of some square matrices. In this section,
before you directly move to finding the inverse of a matrix you have to consider how to develop
the equivalence of two matrices using the elementary row operation techniques. Finally, you will
consider the procedures how to find the inverse of a matrix if it exists.

Elementary Row Operations

In solving a system of linear equation by elimination or addition method, you can transform the
given system to an equivalent system by:
a) interchanging the two equations or
b) Multiplying any equation by a non-zero real number.

Procedure in finding the inverse of a square matrix

One can directly use the definition to find the inverse of a matrix; however you are required to
solve n linear equations with n unknowns. Hence, using the definition is not a wise method to
find the inverse of a matrix especially if the order of the matrix is three or more.

Let’s begin by considering a tabular format where the square matrix A of order n is augmented
with an identity matrix of the same order n as  A | I n , where the two matrices are separated by a
vertical line.
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 1 I n 
Then because AA1 = I and A1 I n  A1 then, we would have  I n | A 1 . We don’t follow this
procedure, because the inverse is not known at this point; we are trying to determine the inverse.
We instead employ a set of permissible elementary row operations on the augmented matrix
A | I n  , to transform A on the left of the vertical line in to an identity matrix ( I n ). As the
identity matrix is formed on the left of the vertical line, the inverse of A is formed on the right
side whenever A has an inverse.

Two basic procedures to find the inverse of a square matrix

1. To set ones first in a column and next zeros (with in a given column)
2. To set zeros first in a matrix and next ones.
Ones first method

8
  3 2
Example 2.7 Find the inverse of the following matrix A    .
1 1
Solution: First augment the given matrix A with the same dimensional unit matrix of order two.
3 2 1 0  R2
 A | I 2    
1 1 0 1 R1
Interchange the first row with the second row:
1 1 0 1 R1
 
 3 2 1 0  R2  3R1
Do not change row one but change the second row by multiplying the first row by –3 and add to
R2 i.e. R2  3R1 .
1 1 0 1  R1  R2
 
 0  1 1  3   R2
Change the second row by multiplying it row by –1 (i.e.  R2 ) and change the first row by
adding to it the second row (i.e. R1  R2 ).
1 0 1  2 
   ( I 2 | A 1 )
 0 1  1 3 
 1  2
Therefore inverse of A is A 1   
1 3
Ones first method: try to set ones first in a column and then zeros of the same column goes
from left to right

Zeros first method

 3 2
Example 2.8 Find the inverse of the following matrix A    .
1 1
Solution: First augment the given matrix A with the same dimensional unit matrix of order two.
3 2 1 0  R1  2 R2
 A | I 2    
1 1 0 1 R2
Change the first row by adding to it –2 multiplied the second row (i.e. R1  2R2 ) and do not
change the second row:
1 0 1  2  R1
 
1 1 0 1  R2  R1
Do not change the first row but change the second row by adding to it –1 multiplied by the first
row (i.e. R2  R1 ).
1 0 1  2 
 
0 1 1 3 
 1  2
Again the inverse of A is A 1   
1 3

9
Zeros first method: try to set zeros first in a column and then zeros of the same column goes
from left to right.
Exercise: Finding the inverse of a matrix
Try to find the inverse for the following matrices.

1. Find the inverse of the following matrices if they exist?

 2 2 3  2  17 11  1 1 2 
     
a) A   1  1 0  b) B    1 11  7  c) C   2 4  3
 0 3 4   0 3  2  3 6  5 
   
2. What do you conclude from question 1-b and 1-c?
 2 7 1
3. Is it possible to find the inverse of the matrix D    ? Why
  3  9 2

2.5. Matrix Representation of System of Linear Equations in Matrix Form

In this section you will consider how to represent system of linear equations using the matrix
notation. Let’s consider a general linear system of m-linear equations with n-unknowns, given
by:
a11 x1  a12 x 2    a1n x n  b1
a x  a x    a x  b
 21 1 22 2 2n n 2
 ……… (2.1)
         
a m1 x1  a m 2 x 2    a mn x n  bm
where aij , bi are given real numbers and x j are the unknowns or variables for i  1,2,3,, m and
j  1,2,3,, n .
A solution of the system of linear equation (2.1) is an n-tuple of real numbers ( x1 , x2 ,, xn ) ,
which satisfies each of the m-equations.
The above system of linear equation can be described using the matrix method, as matrices are a
compact and convenient way of writing down system of linear equations. Let
 a11 a12  a1n 
 
 a 21 a 22  a 2 n 
A called the coefficient matrix,
   
 
 a 
 m1 a m 2  a mn 
 x1 
 
x 
X   2  called the variable matrix and

 
x 
 n

10
  b1 
 
b 
B   2  called the constant term matrix, then (2.1) can be written using these matrices as:

 
b 
 n
AX  B ………(2.2)
Equation (2.2) is called the matrix representation of the system of linear equation (2.1).
Example: Express the following system of linear equation using matrix notation.
3x  2 y  z  5
4x  2 y 1
x  z 4
Solution First we need to attach zero as a coefficient of the missing variable in each equations in
the system.
3x  2 y  z  5
4x  2 y  0z  1
x  0y  z  4
Then the matrix representation will be
3 2 1  x  5
    
4 2 0   y   1 
1 0  1   z   4 
    
Exercise: Matrix representation of system of linear equation
Please represent the following system of linear equation using matrix notion

2 x  4 z  27

1. Find the matrix representation of 3z  4 x  y  2  0
 3 y  6  3 z

2.6. Solving System of Linear Equation Systems using Matrix
From your high school knowledge, you know how to solve a system of linear equations with two
or three variables, which have the form of
a1 x  a 2 y  a3 z  d1
a1 x  a 2 y  d1 
 and b1 x  b2 y  b3 z  d 2 , respectively using either the substitution
b1 x  b2 y  d 2 c x  c y  c z  d
 1 2 3 3

method or elimination (or addition) method. Obviously as the number of variables (or unknowns)
increases, these methods are not applicable. In this section you will deal with ways of solving
system of linear equations using the matrix method: inverse method and Gaussian elimination
method.
Inverse method
The inverse method is one of the methods used to solve a system of linear equation if the number
of equations and numbers of variables are equal (m  n) and the inverse of the coefficient matrix
exists. Therefore for n  n system of linear equations

11
 a11 x1  a12 x 2    a1n x n  b1
a x  a x    a x  b
 21 1 22 2 2n n 2
 ……… (2.3)
       
a n1 x1  a n 2 x 2    a nn x n  bn
Their matrix representation will be:
AX  B ……… (2.4)
Where A the coefficient matrix is square matrix of order n , X is the variable matrix and B is
the constant term matrix.

The procedure in finding the solution of (2.3) or (2.4) using the inverse method is:
Step-1 Change the system of linear equation into matrix form (i.e. AX  B ).
Step-2 Find the inverse of the coefficient matrix (i.e. A 1 ) , and finally
Step-3 Multiply the inverse of coefficient matrix with the constant term matrix, and the resulting
values are the values of the unknown matrix (i.e. X  A 1 B ).
These three steps are handled as follows.
Given AX  B (You can multiply both sides of the equation from the left-
hand side by A 1 )
A1  AX   A1 B (Using associative property)
 1

A AX A B 1
(Since A 1 A  I n )
I n X  A 1 B (Since I n X  X )
X  A 1 B

Example: Given this system of linear equation applying inverse method we can find the
unknown values.
2 x  3 y  4

 x  2y  2
Solution
Step-1 Change it into matrix form
 2 3 x  4
  the coefficient matrix,   the variable matrix and   the constant term matrix. Hence
1 2   y  2
the matrix form is:
 2 3  x   4 
      
 1 2   y   2
Step 2 Find inverse of the coefficient matrix, augment identity matrix order of two on the right
hand side of the coefficient matrix and apply elementary row operation on it to get the inverse.
 2 3 1 0  (1 2) R1 1 3 2 1 2 0  R1  3R2 1 0 2  3
     
1 2 0 1  R2  (1 2) R1  0 1 2  1 2 1 2 R2 0 1 1 2 
 2  3
Hence the inverse is A 1    .
 1 2 

12
Step 3 Multiply the inverse of the coefficient matrix with the constant term matrix to obtain the
solution.
 x   2  3  4   2 
X          
 y    1 2  2   0 
The value for the unknown variables is x  2 and y  0 .

Limitations of inverse method

- It is only used whenever the coefficient matrix is square matrix
- In addition to apply the method the coefficient matrix needs to have an inverse
- It doesn’t differentiate between no solution and infinite solution cases.

The Gaussian elimination method

If a system of linear equations involves m equations and n variable the corresponding matrix
representation will involve m rows and n columns. The Gaussian elimination method developed
by Mathematician Karl F. Gauss (1777-1855) can help to solve such system of linear equations
such as (2.3) or (2.5) which can be either;
a) the number of equations is equal to the number of variables( m = n),
b) the number of equations is greater than the number of variables( m> n), or
c) the number of equations is less than the number of variables(m< n).
Furthermore, you can identify that the system has a unique solution, infinitely many solution or
no solution. Each case will be handled as follows.

a) If the number of equations is equal to the number of variables (“n” by “n” systems)
In solving a system of linear equation the Gaussian elimination method involves the following
steps.
Step-1 Write the system of linear equation into a matrix form ( AX  B )
Step-2 Augments the coefficient matrix with the constant term matrix. ( A | B )
Step-3 Apply elementary row operation on the augmented matrix and transform it to row-
echelon form.
Step-4 Write the equivalent system of the equation and use back substitution to solve for the
unknowns.

Example 2.10 Solve the following system of linear equation

2 x  3 y  4

 x  2y  2
Solution
Step-1 The matrix representation of the system of linear equation is:
 2 3  x   4 
      
1 2   y   2 
 2 3 4
Step-2 Augment the coefficient matrix with the constant term matrix is   .
1 2 2 
Step-3 Apply elementary row operation on the augmented matrix and transform it to row-
echelon form.

13
  2 3 4  (1 2) R1 1 3 2 2  R1 1 3 2 2
  
1 2 2  R  (1 2) R  0 1 2 0  2R
 0 
0 
  2 1   2  1
Step -4The equivalent system of equation is

 3
x  y  2
 2
 y0
From the last equation y  0 .
3
Put y  0 in the first equation, and then you have x  (0)  2 . Hence, x  2 .
2
Therefore, you have a unique solution x  2 and y  0 .

Example 2.11 Solve the system of linear equation

x  y  2

2 x  2 y  4
Solution
Step-1 The matrix representation of the system of linear equation is:
1 1   x   2 
      
 2 2  y   4
1 1 2 

Step-2 Augment matrix is:  2 2 4 
 
Step-3 Apply elementary row operation on the augmented matrix and transform it to row-
echelon form.
1 1 2  R1 1 1 2 
 
 2 2 4  R  2R  0 0 0 
  2 1  
Step-4The equivalent system of equation is
x  y  2

 00
Hence, you have only one equation with two unknowns. Therefore, there are infinitely many
solutions given by:
( x, y) | y  2  x, for any real number x
That is the set of points on the line x  y  2 are solutions for the system of linear equation
given.
Example 2.12 Solve the system of linear equation
x  y  5

x  y  9
Solution
Step-1 The matrix representation of the system of linear equation is:
1 1  x   5 
      
1 1  y   9 
14
 1 1 5 

Step-2 Augment matrix is: 1 1 9 
 
Step-3 Apply elementary row operation on the augmented matrix and transform it to row-
echelon form.
1 1 5  R1 1 1 2 
 
1 1 9  R  R  0 0 4 

  2 1  
There is no possible operation that we can apply in order to change the primary diagonal entry in
the 2nd column without affecting the first column structure. Hence stop the elementary row
operation.
Step -4The equivalent system of the equation is
x  y  2

 04
Hence, you can observe that the second equation is always false. Therefore, the original system
of linear equation has no solution.
Therefore, Gaussian, unlike the inverse method, makes a distinction between unique solution, no
solution and infinite solutions.

Summarizing our results for solving an “n” by “n” system, we start with the augmented matrix (
A | B ), and attempt to transform it into the row-echelon form of the matrix of one of the
following will result:
1. If the row-echelon form of the augmented matrix is of the type shown below then the system
has a unique solution.
1 0 0 10 
 
For example the row echelon form matrix  0 1 0  5  has unique solution
0 0 1 3 
 
2. If the row-echelon form of the augmented matrix is of the type shown below (all elements in a
given row are zero except the constant term column) then, the system has no solution.
1 0 0 3 
 
For example the row echelon form matrix  0 1 0  5  has no solution.
0 0 0 1 
 

3. A matrix in a form different from (1) and (2), indicates that there are infinite 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.
1 0 2 5 
 
For example the row echelon form matrix  0 1 3  3 has infinitely many solutions.
0 0 0 0 
 
Exercise: Using the inverse method and Gaussian elimination method.
Solve the following system of linear equations applying the method specified.

15
  x  2 y  3z  11

1. Solve 3x  2 y  z  1 , using inverse method.
2 x  y  5 z  11


x  y  z  4

2. Solve 5 x  y  7 z  25 , using the Gaussian elimination method.
2 x  y  3 z  8


x  y  z  4

3. Solve 5 x  y  7 z  20 , using The Gaussian elimination method.
 x  y  3z  8


b) If the number of equation greater than the number of variables (m > n)

To solve an m by n system of equations with m > n, you use all the steps for the Gaussian
elimination method. Now you start with the augmented matrix ( A | B ), and attempt to transform
it into the row-echelon form. In this case one of the following three things will result similar to
the above cases of (a):
1. In the row –echelon form of the augmented matrix, if the bottom m  n rows are all zero, you
have a unique solution as the following example indicates:
3x1  2 x 2  x3  23 1 0 0 3 
 x  3x  2 x  26  
 1 2 3  0 1 0  5 
 , its row –echelon form of the matrix is  .
2 x1  x 2  2 x3  10 0 0 1 4
 
4 x1  5 x 2  3x3  49  0 0 0 0
 
The unique solution is x1  3, x2  5, x3  4 , that is (3,5,4) .
2. In the row –echelon form of the augmented matrix, if the bottom m  n rows are all zero
except in the constant column, then it will indicate that there is no solution for the system as
the following example indicates:
2 x1  x 2  30 1 1 2 15 
  
 x1  2 x 2  24 , its row -echelon form of the matrix is  0 1 6 .
4 x  5 x  72  0 0  6
 1 2  
The equivalent system of linear equation is:
 x1  1 2 x 2  15

 x2  6
 0  6

Since the last equation is always false the system has no solution.
3. A matrix in a form different from (1) and (2) indicates that there are an unlimited number of
solutions as the following example indicates:

16
3x1  2 x 2  x3  6 1 0 1 3 2 
6 x  5 x  3x  12  
 1 2 3 0 1 1 0
 , its row –echelon form of the matrix is  .
9 x1  6 x 2  3x3  18 0 0 0 0
 
15 x1  10 x 2  5 x3  30 0 0 0
 0 
The equivalent system of linear equation is:
 1
 x1  x3  2
 3
 x 2  x3  0
where the number of unknowns is greater than the number of equation, hence you have
infinitely many solution given by:
 1 
( x1 , x2 , x3 ) | x1  2  x3 , x2   x3 for any real number x3 
 3 
In each of the above cases the number of equations is greater than the number of variables. You
will observe how the augmented method is stronger and you have either unique, infinitely many,
or no solution.

1. The augmented matrix is

 3 2 1 23  R2 1 3 2 26  R1 1 3 2 26  R1  (3 7) R2
     
1 3 2 26  R1  3 2 1 23  R2  3R1  0  7  5  55  R2  (1 7) R2
 2 1 2 19  R  2 1 2 19  R  2 R  0  5  2  33  R3  (5 7) R2
  3  3 1
 
 4 5 3 49  R  4 5 3 49  R  4 R  0  7  5  55  R4  R2
  4  4 1 

1 0  1 7 17 7  R1  (1 5) R3 1 0 0 3 
   
 0 1 5 7 55 7  R2  (3 11) R3  0 1 0 5 
 0 0 11 7 44 7  (7 11) R  0 0 1 4 .
  3
 
0 0   
 0 0  R4  0 0 0 0 
Therefore, you have a unique solution given by x1  3, x2  5, and x3  4 .

2. The augmented matrix is

 2 1 30  R1 1 2 24  R1 1 2 24  R1  (2 3) R2 1 0 12 
       
1 2 24  R1  2 1 30  R2  2 R1  0  3  18  (1 3) R2 0 1 6 .
 4 5 72       0 0  6
  R3  4 5 72  R3  4 R1  0  3  24  R3  R2  
Therefore, from the last row you can observe hat the system has no solution.

3. The augmented matrix is

3 2 1 6  (1 3) R1 1 23 13 2
   
6 4 3 12  R2  2 R1 0 0 1 0
9 6 3 18  R3  3R2 0 0 0 0
   
15 10 5 30  R  5 R 0 0 
  4 3  0 0

17
 The equivalent system of linear equation is:
 2 1
 x1  x2  x3  2
 3 3
 x3  0
Therefore, you have two equations with three variables and have infinitely many solutions
 2 
given by ( x1 , x2 , x3 ) | x3  0, x1  2  x 2 for any real number x2  .
 3 

c) If the number of equation is less than the number of variables (m < n)

To solve an m by n system of equations with m < n, you use all the steps for the Gaussian
elimination method. You start with the augmented matrix ( A | B ), and attempt to transform it
into the row-echelon form. In this case one of the following two will result similar to the above
case (a) and (b):

1. A row which is all zeros except in the constant columns, indicating that there are no solutions,
or
2. A matrix in a form different from number one above indicating that there are an unlimited
number of solutions.

“Every system of linear equations has exactly one solution, no solution, or infinitely many
solutions.”

Example 2.13 Solve the following systems of linear equations

4 x1  6 x 2  3x3  12
  x1  3x2  x3  6
1.  9 2. 
6 x1  9 x 2  2 x3  20  x1  x2  x3  2
Solution In each of these cases the number of equation is less than the number of variables. You
will observe that you have either no solution or infinitely many solutions.
1. The augmented matrix is

4 6  3 12  1 4 R1 1 3 2  3 4 3 
 
 6 9  9 2 20 R  (3 2) R  0 0 2 

  2 1
0
Therefore, from the last row of the row-echelon form of the matrix you observe that the system
has no solution.

2. The augmented matrix is

 1 3 1 6  R1 1 3 1 6  R1 1 3 1 6
  
  1 1 1 2 R  R  0 4 2 8  (1 4) R

  0 1 1 2 2 
  2 1   2  
The equivalent system of linear equation is:
 x1  3x2  x3  6

 x 2  1 2 x3  2

18
Therefore, you have two equations with three variables and have infinitely many solutions given
 1 1 
by ( x1 , x 2 , x3 ) | x2  2  x3 , x1  x3 for any real number x3  .
 2 2 
Word problems
In order to solve word problems the following four procedures are important to observe
1. Represent the unknown quantities using the English alphabet letters usually x, y, z, or you
can choose the same letters with subscripts like x1 , x2 , x3 , and so on.
2. Translate the quantities from the statement of the problem in to algebraic form and set up a
system of equations.
3. Solve the system of linear equations for the unknown represented by the letter.
4. Check the findings according to the statement in the problem

Example 2.14
1) A firm manufacturing office furniture finds that it has the following variable costs per unit in
dollar. Cabinets
Tables
Chairs
Desks

Mat er ia l  50 20 15 25 
 
Lab o r  30 15 12 15 
O ver he ad  30 15 8 20 

Assume that an order for 5 desks, 6 chairs, 4 tables and 12 cabinets has just been received. What
is the total material, labor, and overhead cost associated with the production of the ordered
items?
Answer: Material cost = $ 730, Labor cost = $ 468, and Overhead cost = $ 512
2. Kebede Carpet Company has an inventory of 1,500 square yards of wool and 1,800 square
yards of nylon to manufacture carpets. 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 yard of nylon.
If the company wants to use all the material in inventory, how many rolls of superior and
how may rolls of quality carpeting should it manufacture?
Answer: 15 superior-grade and 40 quality-grade rolls.
3. Getahun invested a total of Birr 10000 in three different saving accounts. The accounts paid
simple interest at an annual rate of 8%, 9% and 7.5%. Total interest earned for the year was
Birr 845. The amount in the 9% account was twice the amount invested in the 7.5% account.
How much did Getahun invest in each account?
Answer: 1000, 6000, and 3000

Method of solutions to the word problems

1. You can use matrix multiplication to solve the problem.

19
  50 20 15 25 
 
Let the variable cost per unit be V   30 15 12 15  and the number of order made in each
 30 15 8 20 
 
 5
  Desks
 6
item to be R    Chairs
4
  Tables
12 
  Cabinets
Therefore, the total cost (TC ) will be:
 5
 50 20 15 25   50(5)  20(6)  15(4)  25(12)   730
  6     
TC  V .R   30 15 12 15     30(5)  15(6)  12(4)  15(12)    468
 30 15 8 20  4   30(5)  15(6)  8(4)  20(12)   512
 12    
 
Hence, the material cost is $730, the labour cost is $468, and the overhead cost is $512.
2. Let x  the number of rolls of superior grade carpeting manufactured, and
y  the number of rolls of quality grade carpeting manufactured.

The above information can be displayed in the following table.

Superior grade Quality grade Available
carpeting carpeting amount
Wool requirement per roll 20 sq. yards 30 sq. yards 1500 sq. yards
Nylon requirement per roll 40 sq. yards 30 sq. yards 1800 sq. yards

The system of linear equations is:

20 x  30 y  1500

40 x  30 y  1800
The matrix representation is:
 20 30  x  1500 
     
 40 30  y  1800 
To solve the system using the Gaussian method, the augmented matrix is:
 20 30 1500 (1 20) R1 1 3 2 75  R1  (1 20) R2 1 0 15 
  
 40 30 1800 R  20R  0  30 1200 (1 30) R
  0 1 40
  2 1   2  
Therefore, x  15 and y  40 , which means 15 superior grade and 40 quality grade rolls of
carpeting are manufactured.
3. Let x1  be the amount of Birr invested at 8%,
x 2  be the amount of Birr invested at 9%, and
x3  be the amount of Birr invested at 7.5%.
Since the total investment for the year is Birr 10,000, we have x1  x2  x3  10,000
Since total interest earned for the year was Birr 845, we have
0.08x1  0.09x2  0.075x3  845 , and the amount invested in 9%was twice that of the

20
 amount in the 7.5%, we have x2  2x3 , that is x2  2 x3  0 . Hence the system of linear
equations will be:
 x1  x2  x3  10,000

0.08 x1  0.09 x 2  0.075 x3  845
 x 2  2 x3  0

 x1  x 2  x3  10,000

Equivalently, 80 x1  90 x 2  75 x3  845,000
 x 2  2 x3  0

The matrix representation is:
 1 1 1  x1  10,000 
    
 80 90 75  x2    845,000 
 0 1  2  x   0 
  3   
To solve the system using the Gaussian method, the augmented matrix is:
 1 1 1 10,000  R1  1 1 1 10,000  R1  (1 10) R2
   
 80 90 75 845,000  R2  80 R1  0 10  5 45,000  (1 10) R2
 0 1 2   0 1 2 0  R3  (1 10) R2
 0  R3 
 1 0 32 5,500  R1  R3  1 0 0 1,000 
   
 0 1 1 2 4,500  R2  (1 3) R3  0 1 0 6,000 
 0 0  3 2  4500   0 0 1 3,000 
   (2 3) R3  
Therefore, you have a solution x1  1000, x2  6000, and x3  3000 , which shows that Birr
1,000 is invested at 8%, Birr 6,000 is invested at 9%, and Birr 3,000 is invested at 7.5%.

Activity Ten: Solution of system of linear equation using matrices

1. A certain corporation has four factories, each of which manufactures two products. The
number of units of product i produced at factory j in one day is represented by aij in the
 60 40 20 30 
matrix A    . Find the amount of production if the daily production
 30 90 60 70 
increases by 15%?

2. Aster bought a can of milk and a pack of cigarettes for a total of Birr 4.65 excluding tax.
Including the tax the bill was Birr 4.95. If there is 5% sales tax on milk and 8% sales tax on
cigarette, then what was the price of each item?

3. A special food for patients is to be developed from two foods: “Food M” and “Food N”. The
new food is designed so that it contains exactly 31mg of vitamin, 33 mg of vitamin B, and 29
mg of vitamin C. Each kg of food M contains 3 mg of vitamin A, 5 mg of vitamin B and 1
mg of vitamin C. On the other hand, each kg of food N contains 2 mg of vitamin A, 1 mg of
vitamin B and 3 mg of vitamin C. How many kg of each food should be used in the mixture
in order to meet the above requirement?

21
2.7 Markov Chain Analyses(Reading Assignment)
One of the applications of system of linear equations is on the Markov chain, which is a
forecasting model developed by Russian Mathematician called Andrew Markov around 1907. It
is probabilistic model associated with a system that involves change from one state to another in
such a way that we are only able to assign probability other than 0 and 1 to the occurrence of
future state. Hence, the progression of the system through a sequence of states is called Markov
chain orstochastic process.

Markov chains are models which are useful in studying the evolution of certain systems over
repeated trials such as price movement, consumer’s behavior and many other processes in
business. These repeated trials are often successive time periods where the state (outcome
condition) of the systems in any particular time period can’t 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 probabilities of the
system in a particular state at a given time period. We can also talk about the long run or
equilibrium or steady state.
Necessary assumptions of the chain:
1. The system condition (outcome) state in any given period depends on its state in the
preceding period and on the transition probabilities.
2. The transition probabilities are constant overtime.
3. Change in the system will occur once and only once each period. For example if it’s a
week, then it’s only once in a week.
4. The transition period occurs with regularities. For example if we start with days, we use
the day until we reach our end.
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 makes in an
uncertain movement, all that is known is the probability associated with any possible move or
transition. This probability is known as transition probability, symbolized by P  Pij . It is the
likelihood that the system, which is currently in, state i will move to state j in the next period.
From these inputs the model makes two predictions usually expressed as a row matrix.
1. The probabilities of the system being in any state at any given future time period.
2. The long run (or equilibrium) or steady state probabilities.
The set of transition probabilities is necessary for both prediction (time period n, and steady
state), but the initial state is needed for only the first prediction.
Input data prediction(outcome)

Set of transition The steady states or

About the past
probabilities.
the long run states

Currently / initial
About today
The probabilities of the

states. system being in any state at 22

any given time period.
Example 2.15 Currently it is known that 80% of customers shop at store 1 and 20% shop at store
2. While reviewing past data, suppose we find that out of all customer who shopped at store 1 in
a given week, 90% remain loyal for the next week, 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,
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?

Solution There are two states in the problem. Let S1 be the proportion of customers who shop at
store 1, and S 2 be the proportion of customers who shop at store 2.
Since it is known that 80% of customers shop at store 1 and 20% shop at store 2, the initial state
or current state probability matrix (V1 ) for store one and two will be denoted and given by the
row matrix:
S1 S2
V1  (0.8 0.2)

To find the transition probability matrix (P ) from the information you have, the following
structure shows how it is developed.
Next week
Current week

Transition probability matrix is a square

 0.9 0.1 
 S1 S 2  matrix such that each entry indicates the
S1  0.2 0.8  probability of the system moving from a
S2 given state to another state.

 0.9 0.1 
Therefore, the transition probability matrix will be: P   
 0.2 0.8 
Remark:
1) The transition probability matrix is always a square matrix.
2) The sum of the elements in each row in the transition matrix is always 1.
3) You have to be consistent in writing the elements; they are values between 0 and 1.

To find the next state probability matrix (V2 ) , you use the current state matrix multiplied with the
transition probability matrix (P ) given by:
Vn  Vn1 P , for n  2,3,4, ………(2.5)

Where P = is transition probability matrix

Vn = the nth state probability matrix
Vn1 = the ( n  1 )th state probability matrix
Equation (2.5) is the Markov Chain Formula, which enables you to compute the probabilities
of the system being in any state at any given future time period.

a) Hence, the second state probability matrix (that is the market share in the second week) will
be:

23
  0.9 0.1 
V2  V1 P  0.8 0.2   0.8(0.9)  0.2(0.2) 0.8(0.1)  0.2(0.8) 
 0.2 0.8 
S1 S2
 0.72  0.04 0.08  0.16  (0.76 0.24)
This shows that in the second week 76% of the customers shop at store 1 and 24% of the
customers shop at store 2.
Hence, the third state probability matrix (that is in the third week) will be:
 0.9 0.1 
V3  V2 P  0.76 0.24    0.76(0.9)  0.24(0.2) 0.76(0.1)  0.24(0.8) 
 0.2 0.8 
S1 S2
 0.684  0.048 0.076  0.192   (0.732 0.268)
This shows that in the third week 73.2% of the customers shop at store 1 and 26.8% of the
customers shop at store 2. In computing V3 you have used V 2 but not V1 , hence Markov process
has no memory.

b) Long run market share assumption

In the log run the share of the systems is assumed to be constant. Let the share of the system on
S1 S2
the long run be V  (v1 v2 ) , where v1 is the share of the state S1 and v 2 is the share of the
state S 2 . Then we have:
  0.9 0.1  0.9v1  0.2v 2  v1
VP  V v1 v 2    v1 v2  
 equivalent to   0.2 0.8  , that is 0.1v1  0.8v 2  v 2
v1  v 2  1 v  v  1 v  v  1
1 2  1 2

 0.1v1  0.2v 2  0 (1)


 0.1v1  0.2v 2  0 (2)
 v  v  1 (3)
 1 2

Note that equation (1) and (2) are the same, hence you can choose one of them and solve with
equation (3) using any method.
 0.1v1  0.2v2  0

 v1  v2  1
From the last equation you have v 2  1  v1 and put it in the first equation, then
 0.1v1  0.2(1  v1 )  0
 0.3v1  0.2
2 1
Hence, v1  , and v 2 
3 3
Therefore, in the long run 66.7% of the customers will shop at store 1 and 33.3% of the
customers will shop at store 2.

In Prediction of the Long run: - only the transition matrix is required

24
At specified time: - the transition matrix and the initial state matrix are required. Hence, unless
the transition matrix is affected, the long run state will not be affected. Moreover, we cannot
know the number of years, months or weeks to attain the long run state / point but we can know
their share.

Activity Eleven: Application system of linear equation in the Markov Chain Process
1. A division of the Ministry of Public Health had conducted a simple survey on the public
attitude towards smoking. From the results of the survey the department concluded that
currently only 20% of the population smokes cigarette and every month 10% of non-smokers
start smoking whereas 5% of the smokers quit smoking.
Required
a) Write the current and transition matrices
b) What will be the proportion of the non-smokers and smokers in the long run?

2. Among the products of a company, brand A has 40% of the market share. A market
research firm finds that if a person uses brand A, the probability that he/she will be using it
again next year is 30%. On the other hand if a person is not using the product at present, the
probability that he/she will be using it next year is 60%.
Required:
a) Find the transition matrix.
b) Find the percentage of the market share that brand A gets after two years.
c) Want percentage of the market share will be handled by brand A on the long run,
if the transition matrix does not change?

25