CHAPTER 2

GEOMETRIC METHODS
2.1 Graphical Solution Method
For linear optimization problems that have only two variables, it is possible that the entire set
of feasible solutions can be displayed graphically by plotting linear constraints on a graph
paper to locate the best (optimal) solution. The technique used to identify the optimal
solution is called the graphical solution approach for linear optimization problem with two
variables.
The steps of the methods are summarized below:
1. Develop LP model (state the given problem in the mathematical linear optimization
problem)
2. Plot constraints on a graph paper and decide the feasible region.
3. Examine extremes points (corners) of the feasible solution space
i. Determine the coordinates of each corner (extreme) point of the area of feasible
region (or space) where all constraints overlap.
ii. Evaluate the value of the objective function at each extreme point.
iii. Determine the extreme point of the feasible region that has optimum (best)
objective function value.
Example: Use the graphical method to solve the following linear optimization problem
1. Max Z  15x1  10x2
Subject to:
4 x1  6 x2  36
3x1  0 x2  18
0 x1  5x2  20
And x1 , x2  0
2. Min Z  3x1  2 x2
Subject to:
5x1  x2  10
x1  x2  6
x1  4 x2  12
And x1 , x2  0

___________________________________________________________________ 1
Graphical Solution
2.2 CONVEX SET
Definition: A set S  R n is said to be convex if the line segment between any two points in
S lies in S . i.e. for any x1 , x2  S and any  with 0    1 , we have x1  (1   ) x2  S .

Example: 1 Check whether the following sets are convex or not.

a) S  x / x  4

Solution: - Let x1 , x2  S  x1  4 and x2  4 , let   (0,1)

x1  (1   ) x2  x1  (1   ) x2 =  x1  (1   ) x2  4  (1   )4  4

 x1  (1   ) x2  4

 x1  (1   ) x2  S
Therefore, S is convex.
b) S  x / Ax  b, where A  (aij ) m x n and b  R m

Solution: - Let x1 , x2  S  Ax1  b and Ax 2  b , let   (0,1)

A(x1  (1   ) x2 )  A(x1 )  A((1   ) x2 )
  ( Ax1 )  (1   )( Ax 2 )
  (b)  (1   )(b)  b

A(x1  (1   ) x2 )  b  x1  (1   ) x2  S .
Therefore, S is convex.
Theorem:
i) The intersection of convex set is convex.
ii) If K1 and K 2 are convex sets, then K1  K 2 is also convex set.
iii) If a  R and K is convex, then aK is convex.
iv) Let T : x1  x2 be a linear transformation and K is a convex set. Then

T (k )  z  x2 / z  T ( x), x  K  is convex

___________________________________________________________________ 2
Graphical Solution
Proof:
n
(i) Let x , y   S i , where S i is convex i
i 1

n
We want to show that: x  (1   ) y   S i
i 1

 x  S i and y  S i , for all i

 x  (1   ) y  S i , for all i and   (0,1) , (since S i is convex)

n
 x  (1   ) y   S i , for all i ( by definition of intersection )
i 1

n
 S i is convex.
i 1

Therefore, the intersection of any convex set is convex.

(ii) Let x , y  K1  K 2
We want to show that: x  (1   ) y  K1  K 2

 x  x1  x2 , x1  K1 and x2  K 2 , y  y1  y2 , y1  K1 and y 2  K 2
x  (1   ) y   ( x1  x2 )  (1   )( y1  y2 )
= x1  x2  (1   ) y1  (1   ) y2
= x1  (1   ) y1  x2  (1   ) y 2
= z1  z 2  K1  K 2 ( z1  x1  (1   ) y1  K1 , since K1 is convex and
z 2  x2  (1   ) y 2  K 2 , since K 2 is convex)
 x  (1   ) y  K1  K 2
Therefore, the sum of a convex set is convex set.

(iii) Let x , y  aK . Then, x  ax1 and y  ay1 , x1 , y1  K and x1  (1   ) y1  K

We want to show that: x  (1   ) y  aK
x  (1   ) y   (ax1 )  (1   )(ay1 )
= a(x1 )  a((1   ) y1 )
= a(x1  (1   ) y1 )  az , where z  x1  (1   ) y1  K
 x  (1   ) y  aK
Therefore, aK is a convex set.

___________________________________________________________________ 3
Graphical Solution
(iv) Let u , v  T ( K ) . Then there are x, y  K ( i.e. z  x  (1   ) y  K ) such that u  T (x)
and v  T ( y )
We want to show that: u  (1   )v  T ( K )
u  (1   )v  T (x) + (1   )T ( y )
= T (x  (1   ) y ) (since T is a linear transformation)
 T ( z )  T ( K ) , where z  x  (1   ) y  K
u  (1   )v  T ( K )
Therefore, T (K ) is a convex set.
2.3 POLYHEDRON SETS AND EXTREME POINTS
Definition: A point x in a convex set S is called an extreme point (or vertex) of S, if it
does not lie on any line segment joining any two distinct points say x1 and x 2 in S.
Mathematically, a point x in a convex set S is called an extreme point if there does not exist
any pair of distinct points x1 and x 2 in S, such that: x  x1  (1   ) x2 , 0    1 .
Remark:
1. An extreme point is a boundary point of the set. However, not all boundary points of
a convex set are necessary extreme points.
2. Some boundary points may lie between two other points on the same boundary, since
strict inequalities are imposed on  ( 0    1 ). Therefore, an extreme point cannot
be between any two points of the convex set.
3.
x4
x5

x6
x3
x1

x2

Extreme and non extreme points

___________________________________________________________________ 4
Graphical Solution
Example: x1 , x 2 , x3 , x 4 and x5 are extreme points and x 6 are non-extreme points.

Definition: If the set S  x1 , x2 ,  , xn  consists of finite number of points, then convex
hull of S is called the convex polyhedron. In other words, a set of all convex combinations of
a finite number of points is called the convex polyhedron spanned by these points.
n n 
Symbolically, it may be expressed as C ( S )   j j  x ,  j  0 ,   j  1
 j 1 j 1 
Definition: The convex hull of S   n is the intersection of all convex sets containing S .
2.4 THE CORNER POINT THEOREM
Theorem: If the feasible region is bounded, then the objective function has both a maximum
and a minimum value and each occur at one or more corner points. If the feasible region is
unbounded, the objective function may not have a maximum or a minimum value. But if a
maximum or minimum value exists, it will occur at one or more corner points.

___________________________________________________________________ 5
Graphical Solution