Linear Algebra

Course Outline
Course Outline
Course Outline
Course Outline
Course Outline
Chapter One: Linear Algebra
I. Introduction
• Linear Algebra
• Permits expression of a complicated system of equations in a succinct, simplified way
• Provides a shorthand method to determine whether a solution exists before it is
attempted,
• Furnishes the means of solving the equation system
• Can be applied to system of linear equations ( Many economic relationships can be
approximated by linear equations, if not they can be converted to linear)
I. Introduction
I. Introduction: Definition and Terms
• A matrix is a rectangular array of numbers, parameters, or variables, each of which
has a carefully ordered place within the matrix.
• The numbers (parameters, or variables) are referred to as elements of the matrix. The
numbers in a horizontal line are called rows; the numbers in a vertical line are called
columns.
• The number of rows r and columns c defines the dimensions of the matrix (rxc),
which is read “r by c”. The row number always precedes the column number. In a
square matrix, the number of rows equals the number of columns (that is, r =c).
• If the matrix is composed of a single column, such that its dimensions are r*1, it is a
column vector and if the matrix is a single row, with dimensions 1xc, it is a row
vector.
• A matrix which converts the rows of A to columns and the columns of A to rows is
called the transpose of A and is designated by A’ or AT
• The transpose of A is
I. Introduction: Addition and Subtraction of Matrices
.
I Introduction: Scalar Multiplication
.
I Introduction: Vector Multiplication
.
I Introduction: Vector Multiplication
.I Introduction: Multiplication Matrices
• The matrices should be conformable (Column of the first matrix or lead
matrix should be equal to number of rows in the second or lag matrix)
• Each row vector in the lead matrix is then multiplied by each column vector
of the lag matrix
• The row-column products, called inner products or dot products, are then
used as elements in the formation of the product matrix, such that each
element cij of the product matrix C is a scalar derived from the multiplication
of the ith row of the lead matrix and the jth column of the lag matrix.
I. Commutative, Associative, and Distributive Laws in Matrix Algebra
I. Identity and Null Matrices
• Identity matrix (I) is square matrix which has 1 for every element on the
principal diagonal from left to right and 0 every where else.
• When a subscript is used, as in In , n denotes the dimensions of the
matrix(nxn).
• Identity matrix is like the number 1 in algebra as multiplication of a matrix by
an identity matrix leaves the original matrix unchanged (AI=IA=A)
• Multiplication of an identity matrix by itself leaves the identity matrix
unchanged: IxI=I2=I
• Any matrix which A=A- is a symmetric matrix
• A symmetric matrix for which AxA=A is an idempotent matrix. The identity
matrix is symmetric and idempotent.
• A null matrix is composed of all 0s and can be any dimension
• It is not necessarily a square matrix
• Addition or subtraction of the null matrix leaves the original matrix unchanged
• Multiplication of a null matrix produces a null matrix.
Matrix Expression of a System of Linear Equations
• Matrix algebra allows concise expression of system of linear equations
• A is the coefficient matrix, X is the solution vector and B is vector of constant

terms
• X and B will always be column vectors
Diagonal Matrices
• These matrices have entries on the diagonal and either
above or below the diagonal, while the other part of the
matrix consists of zeros.
• A is an upper triangular matrix and B is a lower triangular
matrix
Determinant of A matrix
• The Determinants |A| of a 2x2 matrix ( second-order determinant) is
derived by taking the product of the two elements on the principal
diagonal and subtracting from it the product of the two elements off
the principal diagonal.
𝑎11 𝑎12
• Given a 2x2 matrix
𝑎21 𝑎22
• The determinant
• The determinant is a single number or scalar and is found only for

square matrices.
• If the determinant of a matrix is equal to zero, the determinant is said
to vanish and the matrix is termed singular
• A singular matrix is one in which there exists linear dependence
between at least two rows or columns
• If |A|≠0, matrix A is nonsingular and all its rows and columns are
linearly independent
• If linear dependence exists in a system of equations, the system as a
whole will have an infinite number of possible solutions ( no unique
solution) ( that is if |A|=0 )
• If |A|≠0, the matrix is nonsingular and there is no linear dependence
among the equations. A unique solution can be found
Determinant of A matrix: Rank(𝝆)
• The rank of a matrix is defined as the maximum number of linearly
independent rows or columns in the matrix.
• The rank of a matrix also allows for a simple test of linear dependence which
follows immediately.
• Assuming a square matrix of order n,
• Example
Third order Determinants
is called a third-order determinant and is the summation of three products.

• To derive the three products:
Third order Determinants
Next slides
Third order determinant
Minors and Cofactors
• The elements of a matrix remaining after the deletion process described above
form a sub-determinant of the matrix called a minor.
• The minor |Mij| is the determinant of a submatrix formed by deleting ith row
and jth column of the matrix.
• Where |M11| is the minor of a11, |M12| is the minor of a12 and |M13| is the
determinant of a13.
• Thus, the determinant of the above the previous 3x3 matrix equals
• A cofactor is a minor with a prescribed sign.
• The rule for the sign of the cofactor is
• Example: The cofactors (1) |C11| , (2) |C12| , and (3) |C13| for the matrix
Laplace Expansion and Higher-order Determinants
• Laplace expansion is a method for evaluating determinants in terms of
cofactors.
• It thus simplifies matters by permitting higher-order determinants to
be established in terms of lower-order determinants.
• Laplace expansion of a third-order determinant can be expressed as
• Laplace expansion permits evaluation of a determinant along any row or
column.
• Selection of a row or column with more zeros than others simplifies evaluation
of the determinant by eliminating terms.
• Laplace expansion also serves as the basis for evaluating determinants of orders
higher than three.
• Example
• Laplace expansion for a fourth-order determinant is
• where the cofactors are third-order subdeterminants which in turn can be

reduced to second-order subdeterminants, as above.
• Fifth-order determinants and higher are treated in similar fashion
Properties of a Determinant
• The following seven properties of determinants provide the ways in which a
matrix can be manipulated to simplify its elements or reduce part of them to
zero, before evaluating the determinant:
Properties of a Determinant
Partitioning A matrix
• Splitting up into smaller matrices
Partitioning A matrix
•.
Operation with Partitioned/block matrices
•.
•.
•.
This is the definition of matrix multiplication: the columns of AB are A multiplied

by the columns of B
•.
Another understanding of matrix multiplication: the rows of AB are the rows of A

multiplied by B.
•.
Cofactor and Adjoint Matrices
Cofactor and Adjoint Matrices
Inverse of A matrix
Inverse of A matrix
Inverse of A matrix
Solving Simultaneous Equations using Matrix
• An inverse matrix can be used to solve matrix equations.
• If
•
Solving Simultaneous Equations using Matrix
•
CRAMER’S RULE FOR MATRIX SOLUTIONS
• Cramer’s rule provides a simplified method of solving
a system of equations through the use of determinants.
|𝐴𝑖 |
• It states 𝑥ҧ𝑖 =
|𝐴|
• where xi is the ith unknown variable in a series of equations,
|A| is the determinant of the coefficient matrix, and |𝐴𝑖 | is
the determinant of a special matrix formed from the original
coefficient matrix by replacing the column of coefficients of xi
with the column vector of constants.
•
•
Gauss Elimination Method
• The idea is to add multiples of one equation to the others in order
to eliminate a variable and to continue this process until one
variable is left.
• Once this final variable is determined, its value is substituted back
into the other equations in order to evaluate the remaining
unknowns.
• This method, characterized by step‐by‐step elimination of the
variables, is called Gaussian elimination.
•
•
•
•
•
•
•
•
• Since the coefficient matrix has been transformed into echelon form, the
“forward” part of Gaussian elimination is complete
• What remains now is to use the third row to evaluate the third unknown, then to
back‐substitute into the second row to evaluate the second unknown, and,
finally, to back‐substitute into the first row to evaluate the first unknown
•
•.
•
•
•
Gauss Jordan Method
• Reduce ( eliminate entirely) the computations involved in
back-substitution above by performing additional row
operations to transform the matrix from echelon to reduced
echelon form.
• A matrix is in reduced echelon form when, in addition to being
in echelon form, each column that contains a non-zero entry(
usually made to be 1) has zeros not just below that entry but
also above that entry.
Gauss Jordan Method
•
Gauss Jordan Method
•
Gauss Jordan Method
•
Gauss Jordan Method
•
Gauss Jordan Method
•
Gauss Jordan Method
•
Gauss Jordan Method
•
Gauss Jordan Method
•
Gauss Jordan Method
•
Homogeneous System of Linear Equations
• The constant term in every equation is equal to zero(0); no equation in
such systems has a constant term in it.
• A system with n unknowns
• A homogeneous linear system may have one or infinitely many

solutions. But it has at least one solution always.
• A homogeneous system may have two types of solutions: trivial solutions and
nontrivial solutions.
• Since there is no constant term present in the homogeneous systems, (x₁, x₂, ...,
xₙ) = (0, 0, ..., 0) is obviously a solution to the system and is called the trivial
solution (the most obvious solution).
• For example, the system formed by three equations x + y + z = 0, y - z = 0, and
x + 2y = 0 has the trivial solution (x, y, z) = (0, 0, 0). But it may (or may not)
have other solutions than the trivial solutions that are called nontrivial
solutions.
• We can find them using the matrix method and applying row operations.
THE JACOBIAN
• Jacobian determinant permits testing for functional dependence, both
linear and non-linear.
• A Jacobian determinant |J| is composed of all the first-order partial
derivatives of a system of equations, arranged in ordered sequence.
• Given
THE JACOBIAN
• Elements of each row are the partial derivatives of one function yi with
respect to each of the independent variables x1, x2, x3, and the elements of
each column are the partial derivatives of each of the functions y1, y2, y3
with respect to one of the independent variables xj.
• If |J|=0, the equations are functionally dependent
• If |J|≠ 0, the equations are functionally independent

THE JACOBIAN
THE HESSIAN
• A convenient test for this second-order condition is the Hessian

• A Hessian |H| is a determinant composed of all the second-order partial
derivatives, with the second-order direct partials on the principal diagonal
and the second-order cross partials off the principal diagonal.
THE HESSIAN
• Thus,
THE HESSIAN
THE HESSIAN
• Example
Higher Order HESSIANS
conditions for a relative maximum are met

THE BORDERED HESSIAN FOR CONSTRAINED OPTIMIZATION
• To optimize a function f(x,y) subject to a constraint g(x,y), a new
function could be formed F(x,y, 𝜆) = f x, y + 𝜆 k − g x, y , where
the first order conditions are Fx=Fy=F λ=0
THE BORDERED HESSIAN FOR CONSTRAINED OPTIMIZATION
•
ഥ 3 <0, 𝑯
𝑯 ഥ 4> 0, etc., the bordered Hessian is negative definite, and a negative definite Hessian always
meets the sufficient condition for a relative maximum.
Eigenvalues and Eigenvectors
• Eigenvalue is a number λ which when subtracted from the
diagonal of a square matrix A results in the matrix to be
singular.
• We say that a matrix is singular when the inverse of the
matrix does not exist. This means the determinant is 0 and
the rank is less than n.
• Consider the following matrix
• The matrix is singular as all columns are exactly the same now and
the rank of the matrix thus is less than 3.
• Therefore, the number 2 is an eigenvalue of this matrix.
• Example :
Eigenvectors
• Eigenvectors are vectors that correspond to a certain eigenvalue and
can be found by solving the equation:
Eigenvectors
•
Eigenvectors
•
Economic Applications
•
•
•
•
•
•
•
•
•
•
•
•
INPUT-OUTPUT MODEL(LEONTIEF MODEL)
• Input-output analysis is a type of economic model that
describes the interdependent relationships between
industrial sectors within an economy.
• It shows how the outputs of one sector flow into another
sector as inputs.
• I-O analysis is a technique which was invented by Professor
Wassily W. Leontief in 1951 won Nobel prize in Economics in
1973 .
INPUT-OUTPUT MODEL: Features of the Model
INPUT-OUTPUT MODEL: Assumptions
INPUT-OUTPUT MODEL
• Suppose an economics system that consists of four producing sectors( three
inter-industry sectors: agriculture, manufacturing, service, and, and one final
demand sector that is the household sector).
• The table below provides a simplified example of such an economy.
INPUT-OUTPUT MODEL
• The data in any row show the distribution of output to
various sectors and uses
• The data in any column indicate the source of inputs needed
for production.
• For example, reading across the first row, altogether the
agricultural output is valued at birr 300. Of his total, 100
birr worth of goods goes to final consumption; the remaining
output from agriculture goes as inputs-50 birr to itself, 150
birr to manufacturing and no output to service sectors.
INPUT-OUTPUT MODEL
• An I-O table is based up on “transaction matrix”.
• A transaction matrix shows how the total output of one industry is distributed
to all other industries as inputs and to the final demand sector.
• Now we can convert the above input output table in the form of “transaction
matrix” as given below.
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
• We can add the entries in the transaction matrix row wise to get the
output of each industry. Thus,
INPUT-OUTPUT MODEL
• Now we can construct a table indicating the input requirement for the
production of one birr worth of output. For example, to produce br 300 worth
of agricultural goods, agriculture requires br 50 worth of agricultural goods, br
100 worth of manufacturing and br 150 worth of labor.
• Because of the assumption of fixed proportion, producing br 1 worth of
agricultural output requires
INPUT-OUTPUT MODEL
• By applying the same analysis to other sectors i.e. manufacturing and service, we
have to find what worth of agricultural, manufacturing and service are required to
produce br 1 worth of manufacturing and service output.
INPUT-OUTPUT MODEL
• For our four-sector economy, the technical coefficients would be:
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL
INPUT-OUTPUT MODEL: Determination Of Employment Levels
Linear Programming
• Definition of LP
• Formulate LPP
• Solve LPP using Graphic Method
• Solve LPP using Simplex Method
• Exercise Application of LP in different areas of
Economics
Definition of LP
• LP is a composite of two words: Linear and
Programming
• Linear mean all mathematical functions in the model
are required to be linear functions
• Programming is planning aspect of activities
• LP is a tool or technique by which one can select the
optimal solution of problems like distribution of
commodities, allocation of machine, scheduling
transport services, etc….
Steps in LP
1. Identify the Objective function as a linear function of its variables and
state all the limitations on resources as linear equations or inequalities
2. Use mathematical techniques to find all possible sets of values of the
variables or unknowns, satisfying the constraints
3. Select the particular set of values of variables obtained in step two that
leads to the achievement of the objective function
▪ The result of step one is called linear programming problem.
▪ The set of solutions in step two are known as the set of feasible
solutions
▪ The solution finally selected in step three is called optimum(best)
solution of the LPP.
Steps in LP
• Therefore, a typical LPP has three parts:
• The objective function
• The constraints
• The non-negative principle(constraint)
The Objective Function
• Objective lies on two extremities: either maximizing or minimizing
something ( maximize profit, sales, equity, efficiency, production… or
wants to minimize cost, time, pollution, etc.
• The objective function can be stated as
The Constraints
• Constraints are situations under which the objective function should be
maximized or minimized
• It answers questions like how much money, time, resource, technical
capacity, etc are available to achieve the objective function
• A constraint is formulated in linear form and the constraint is expressed
in either <, ≤, ≥, =, or >.
• It can be formulated as
The Constraints
The Non-Constraint
• The non-negativity constraint is a principle which prohibits that any one of
the variables be negative
• In economics, generally, and in linear programming specifically, negative
numbers are rarely used. This is because using negative numbers may be
misleading
• For example, what does it mean to produce -20 quintals of teff or -25
quintals of maize?
• Any producer cannot produce negative output from use of inputs
• Generally, the decision (choice) variables should be non-negative.
• In linear programming the negative aspect of variables should be given as:
Assumptions of Linear Programming
• Proportionality: in both the objective function and the
constraints
• If you want to double the output, we need simply to double the
resources
• Additivity:
It states that the sum of resources used by different activities must be equal to
the total quantity of resources used by each activity for all resources
individually and collectively. In other words, interaction among the activities of
resources does not exist.
• Divisibility: This assumption implies that the solution need not be in
whole numbers(integers). Instead, they are divisible and may take any
fractional value.
Assumptions of Linear Programming
• Certainty:
We assume that the conditions of certainty exist. That is , coefficients
in the objective function and constraints are completely known(
deterministic) and do not change during the period being studies. For
example: profit per unit of product and amount of resources available
are fixed during the planning period.
• Finiteness: an optimum solution cannot be computed in the
situations where there are infinite number of alternative activities and
resource restrictions
• Optimality: In LP, the maximum profit solution or the minimum
cost solution always occur at the corner point of the set of feasible
solutions
Formulations of of Linear Programming Problem
• Identify the objective function
• Identify the variables
• State the feasible alternatives. Non-negative constraint of
variables should be imposed
• Express the constraints in terms of the variables
• Express the objective function quantitatively in terms of
the variables
• The standard form of the model (LPP) can be stated as
• The function being maximized is called the objective function
and the restrictions are called the constraints.
Methods of Solving Linear Programming Problems
•The graphic and
•The simplex (algebraic) methods
Methods of Solving LPP: Graphic Method
• LPP involving only variables can be solved easily by graphic method.
• It entails the following procedures
• Graph the constraints
• Identify the solution space
• Locate the solution points
• Identify the corner points from the solution space. This is because the
optimal solution of a linear programming problem is always found at the
corner point of the solution space under normal circumstances
•Determine the optimal solution. A solution that optimizes the objective
function.
• After graphing the constraints, the next step is to locate the corner
points since optimal solution is obtained at one of the corner point
feasible basic solution
• The optimum solution is multiple which lies on the ray between (3,
6) to (6, 3). That means all points lying on this ray are optimal. For
example, the point (4, 5) and (5, 4) are the coordinates that belong
to this ray and the profit is to be 990. Therefore, the linear
programming is a multiple optimal solution, i.e. there is more than
one optimal solution.
Unbounded solution: - This is a situation that occurs whenever
a maximization LPP has unbounded solution space. No finite
solution can be determined. This is because one or more of the
variables can be increased indefinitely without violating the
feasibility.
Mostly this situation occurs when the assumption of finiteness
is violating.
Example: - solve the following linear programming problem
using graphic method.
Unbounded solution
The corner points provide the minimum profit here not

the maximum profit. The profit is increasing indefinitely
without any bound. In this case the linear programming
is unbounded solution space linear programming
• Infeasible Solution: - This indicates that there may be a solution
which can satisfy each constraint but there is no feasible solution.
This is because there is no point that satisfies all the constraints of
the LPP. In such case we say that the LPP has no feasible solution.
The Simplex Method
• Graphic method will not be used when the number of variables is more than
two
• It is not possible to depict graphically in two-dimensional plane
• Alternative comprehensive method of solving LPP=> Simplex/algorithm
• Algorithm is a process where a systematic procedure is repeated (iterated) over
and over again until the desired result is obtained.
Maximization of Simplex Method
Setting up the Simplex Method
Procedures of Simplex Method
From the above table we can read the initial solution for the problem as S1 = 48, S2 = 15, S3= 10 and others zero with zero
profit.
• Step 3: Test the solution for optimality. The test for optimality can be
formalized in terms of the sign of the members of the index row. The initial
solution is optimum if all the elements of index row are non-negative. If at least
one of the elements of the index row is negative, the given solution is not
optimal and goes to the next step.
Procedures of Simplex Method: Minimization
• In minimization case even though the basic process of simplex is the same, there
are some arrangements.
• One of those arrangements is the use of artificial variables
• Let’s discuss about artificial variables before going through minimization
2
• There are some steps which are different in minimization from maximization
are: Test for optimality and means of selecting entry variable.
• For minimization test for optimality is carried out until all elements of net
evaluation row or index row are non-positive.
• That means, a given solution is not optimal if at least one element of the index
row is positive.
• The variable which has the largest positive value is selected as entry variable
and the column containing this variable is a pivot column.
• But there is no any difference in selecting the leaving variable and pivot number
in minimization from that of maximization.

Linear Algebra

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Linear Algebra

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Linear Algebra

Uploaded by

Copyright:

Available Formats

Course Outline

• A is the coefficient matrix, X is the solution vector and B is vector of constant

• The determinant is a single number or scalar and is found only for

is called a third-order determinant and is the summation of three products.

• where the cofactors are third-order subdeterminants which in turn can be

This is the definition of matrix multiplication: the columns of AB are A multiplied

Another understanding of matrix multiplication: the rows of AB are the rows of A

• A homogeneous linear system may have one or infinitely many

• If |J|=0, the equations are functionally dependent

• If |J|≠ 0, the equations are functionally independent

• A convenient test for this second-order condition is the Hessian

conditions for a relative maximum are met

The corner points provide the minimum profit here not

You might also like