MODM: Goal Programming

Prof. Pankaj Dutta

SJM School of Management
IIT Bombay
Multi-Objective Decision Making (MODM)
 Usually, we consider decision models that optimizes a single objective,
such as profit maximization or cost minimization. But, in most situations,
we encounter in business and life, however, more than one relevant
objective exists.
 For example, in optimal investment portfolio, you want to maximize expected return,
but you also want to minimize risk. How do you reconcile these conflicting objectives?
 In this module, though we will emphasis mainly on MCDM, but at the
same time, the concerns/issues related to MODM, to some extent, will also
be addressed.
 However, while addressing the MODM problems, we will restrict our
discussions on

 Pareto Optimality (Basic)

 Goal Programming
 Interactive Fuzzy Decision Making
 Analytic Hierarchy Process
Pareto optimal solution
 Researchers study multi-objective optimization problems from
different viewpoints and, thus, there exist different solution
philosophies and goals when setting and solving them. The goal may
be to find a representative set of Pareto optimal solutions, and/or
quantify the trade-offs in satisfying the different objectives, and/or
finding a single solution that satisfies the subjective preferences of a
human decision maker .
 For a nontrivial multi-objective optimization problem, there does
not exist a single solution that simultaneously optimizes each
objective. In that case, the objective functions are said to be
conflicting, and there exists a (possibly infinite number of) Pareto
optimal solutions.
Pareto optimality
 A solution is called non-dominated, Pareto optimal, Pareto efficient
or non-inferior, if none of the objective functions can be improved
in value without degrading some of the other objective values.
Without additional subjective preference information, all Pareto
optimal solutions are considered equally good.
Or
 A solution is Pareto optimal if there exists no feasible solution for
which an improvement in one objective does not lead to a
simultaneous degradation in one (or more) of the other objectives.
That solution is a non-dominated solution.
 In Economics, Pareto optimum is the conditions under which the
state of economic efficiency (where no one can be made better off
by making someone worse off) occurs.
 Pareto optimality
 If it is possible to make at least one person better off in moving from state A to
state B without making anyone else worse off, state B is ranked higher by
society than state B.
 Given an initial allocation of goods among a set of individuals, a change to a
different allocation that makes at least one individual better off without
making any other individual worse off is called a Pareto improvement. An
allocation is defined as "Pareto efficient" or "Pareto optimal" when no further
Pareto improvements can be made.
 For example, suppose there are two consumers A & B and only one resource
X. Suppose X is equal to 20. Let us assume that the resource has to be
distributed equally between A and B and thus can be distributed in the
following way: (1,1), (2,2), (3,3), (4,4), (5,5), (6,6), (7,7), (8,8), (9,9), (10,10).
At point (10,10) all resources have been exhausted. No further distribution is
possible - if redistribution continues, it will lead to a position (11,9) or (9,11)
that makes one better off and the other worse off. Hence, point (10,10) is
Pareto optimal; no further Pareto improvements can be made.
 Now, if X is equal to 19, then what is your Pareto point?
Goal Programming
 What is Goal Programming?
 Basic Difference between GP and AHP
 Goal Programming Approach and Steps
 Goal Programming Constraints
 Graphical Interpretation of GP
 Weighted Methods
 Preemptive Methods
 Illustrative Examples..
What is Goal Programming?

 Mathematical model similar to Linear Programming, however

it allows for multiple goals to be satisfied at the same time.
 Allows for the multiple goals to be prioritized and weighted to
account for the DM’s utility for meeting the various goals.

 Examples..
Basic difference between GP and AHP
 GP answers “how much?”, whereas AHP answers “which
one?”
 GP: In many situations, a company wants to achieve several
objectives. Given limited resources, it may prove impossible
to meet all objectives simultaneously. If the company can
prioritize its objectives, then Goal Programming can be used
to make good decisions.
 AHP: Method for ranking decision alternatives and selecting
the best one when the decision maker has multiple
objectives, or criteria
Multi-Objective Decision Making: Goal
Programming

 In some situations, a decision maker may face

multiple objectives, and there may be no point in an
LP’s feasible region satisfying all objectives.
 In such a case, how can the decision maker choose a
satisfactory decision?
 Goal programming is one technique that can be
used.
Goal Programming Approach
The basic approach of goal programming is:
 To establish a specific numeric goal for each of
the objectives.
 Formulate an objective function for each
objective
 Seek a solution that minimizes the weighted sum
of deviations of these objective functions from
their respective goals.
Types of Goals
 A lower, one-sided goal sets a lower limit.
 An upper, one-sided goal sets an upper limit.
 A two-sided goal sets a specific target that we do
not want to miss on either side.
Main Objective is to: Minimizes the sum of the
weighted deviations from the target values.
Each goal will minimize the undesired deviation.
Goal Programming Steps
Step1. Define decision variables
Step2. Define Deviational Variable for each goal
Step3. Formulate Constraint Equations
 Economic constraints
 Goal constraints
Step4. Formulate Objective Function
Goal Programming Constraints

 Decision Variables are the same as those in LP formulations

(represent products, hours worked)

 Deviational Variables represent overachieving or

underachieving the desired level of each goal
 d+ Represents overachieving level of the goal
 d- Represents underachieving level of the goal
Goal Programming Constraints
 Economic Constraints
Stated as <=, >=, or =
Linear (stated in terms of decision variables)
Example: 3x + 2y <= 50 hours

 Goal Constraints
General form of goal constraint:

Decision Desired Goal

Variables
- d+ + d - = Level
 Let’s define :
di+ : amount by which we numerically exceed the ith
goal
di- : amount by which we are numerically below the
ith goal.
di+, di- : deviational variables

 Since both under and over-achievement of a goal cannot

be achieved simultaneously, either one or both of these
deviational variables will be equal to zero.
 If the goal is achieved exactly, both di+ and di- will be
zero.
Example 1
 Consider a LP problem.
Max Z = 40 x1+ 35 x2
St
2x1+ 3x2 < 60
4x1+ 3x2 < 96
x1 , x2 >0

Now, suppose the manager has set the goal to achieve a

profit of Rs 1400.
In this case, 40 x1+ 35 x2 > 1400 is the goal constraints.
Example 1
 The GP is
Minimize Z = d-
St
2x1+ 3x2 < 60
4x1+ 3x2 < 96
40 x1+ 35 x2 +d- - d+ = 1400
x1 , x2, d- , d+ >0
GP: Example 2: ABC Ltd.
 ABC is a growth oriented firm which establishes monthly
performance goals for its sales force.
 ABC determines that the sales force has a maximum available
hours per month for visits of 640 hours.
 Further, it is estimated that each visit to a potential new client
requires 3 hours and each visit to a current client requires 2
hours.
 ABC establishes two goals for the coming month:
 Contact at least 200 current clients
 Contact at least 120 new clients
 Overachieving either goal will not be penalized
GP: Example 2

 Steps Required:
1. Define the decision variables
2. Define the goals and deviational variables
3. Formulate the GP Model’s Parameters:
 Economic Constraints
 Goal Constraints
 Objective Function
4. Solve the GP using the graphical approach
GP: Example 2
 Step 1: Define the decision variables:
X1 = the number of current clients visited
X2 = the number of new clients visited
 Step 2: Define the goals:
Goal 1 – Contact 200 current clients
Goal 2 – Contact 120 new clients

 Step 3: Define the deviational variables

d1+ = the number of current clients visited in excess of the goal of
200
d1- = the number of current clients visited less than the goal of 200
d2+ = the number of new clients visited in excess of the goal of 120
d2- = the number of new clients visited less than the goal of 120
GP: Example 2
 Formulate the GP Model:
 Economic Constraints:
2X1 + 3X2 <= 640 (note: can be <, =, >)
X1, X2 => 0
d1+, d1-, d2+, d2- => 0
 Goal Constraints:
 Current Clients: X1 + d1- - d1+ = 200
 New Clients: X2 + d2- - d2+ = 120 Must be =

 Objective Function:
 Minimize Weighted Deviations
i.e., Minimize Z = d1- + d2-
GP: Example 2
 Complete formulation:
Minimize Z = d1- + d2-
Subject to:
2X1 + 3X2 <= 640
X1 + d1- - d1+ = 200
X2 + d2- - d2+ = 120
X1, X2 => 0
d1+, d1-, d2+, d2- => 0

 Graph constraint???
2X1 + 3X2 = 640
If X1 = 0, X2 = 213
If X2 = 0, X1 = 320
Plot points (0, 213) and (320, 0)
Graphical Solution: Example 2

X2
(0,213)
200

150

100

50
(320,0)
X1
0 50 100 150 200 250 300 350
Graphical Solution: Example 2

 Graph deviation lines

 X1 + d1- - d1+ = 200 (Goal 1)
 X2 + d2- - d2+ = 120 (Goal 2)

 Plot lines for X1 = 200, X2 = 120

Graphical Solution: Example 2

X2
Goal 1
(0,213)
200
d1-
d1+
150 d2+
Goal 2
(140,120)
100 d2-
(200,80)
50
(320,0)
X1
0 50 100 150 200 250 300 350
Graphical Solution: Example 2
 Want to Minimize d1- + d2-
 So we evaluate each of the candidate solution
points:

For point (140, 120) For point (200, 80)

d1- = 60 and d2- = 0 d1- = 0 and d2- = 40
Z = 60 + 0 = 60 Z = 0 + 40 = 40

Contact at least 200 current clients

Contact at least 120 new clients
Goal Programming Solution: Example 2

 X1 = 200 Goal 1 achieved

 X2 = 80 Goal 2 not achieved
 d1+ = 0 d2+ = 0
 d1- = 0 d2- = 40

 Z = 40
Example 3: An electric company
 An electric company produces two products: Grinders and ceiling
fans. Both the products require a two-step production process
involving wiring and assembly. It takes 2 hours to wire each
grinder and 3 hours to wire a ceiling fan. Final assembly of the
grinders and fans requires 6 and 5 hours, respectively.
The production capability is such that only 30 hours of assembly
time are available. At least 12 hours of wiring have to be utilized
due to certain rental constraints. If each grinder produced gives the
firm Rs. 700 and each fan Rs. 600, the LPP is

Max Z = 7x1+ 6x2.

St
2x1+ 3x2 ≥ 12 (wiring)
6x1+ 5x2 ≤ 30 (assembly)
x1, x2 > 0
Example 3
 The firm has the following Goals:
 Goal 1: to produce as much profit above Rs 3000
during the production period.
 Goal 2: to utilize at least 12 hrs wiring.
 Goal 3: to avoid overtime in the assembly
department.
 Goal 4: to meet a contract requirement to produce
at least seven ceiling fans

 Formulate it as a GP.
Example 3
 The Goals can be stated as :
7x1+ 6x2 > 30
2x1+ 3x2 > 12 (wiring)
6x1+ 5x2 < 30 (assembly)
x2 > 7
x1 > 0

 Let’s define :
di+ : amount by which we numerically exceed the ith goal
di- : amount by which we are numerically below the ith goal.
di+, di- : deviational variables
Example 3
 The goal programming problem can be
formulated as:
Min Z = d1- + d2- +d3+ + d4- .
St
7x1+ 6x2 –d1+ + d1- = 30
2x1+ 3x2 –d2+ + d2- =12
6x1+ 5x2 –d3+ + d3- =30
x2 –d4+ + d4- = 7
x1, x2 , di+, di- >0  i =1,2,3,4.
Goal Programming Algorithms

 Weighted method: A single objective function is

formed as the weighted sum of the functions representing
the goals of the problem.

 Preemptive method: It starts by prioritizing the

goals in order of importance. That is, satisfy goals in a
priority sequence so that second-priority goals are pursued
without reducing the first-priority goals, etc.
GP Algorithms: Special Notes

 If the deviational variables in the objective function have no

priorities, we can continue to use the standard simplex
algorithm to solve the problem.
 If priority index of the deviational variables are in the
objective function, we can not use the standard simplex
algorithm to solve the problem. Use a modified GA simplex
algorithm.
Weighted method
 Suppose that the goal programming model has n
goals and that the ith goal is given as
Minimize Gi = 1,2,…,n
 The combined objective function used in the weights
method is then defined as
Minimize z = w1G1+ w2G2+ …+wnGn
 The parameters wi, i = 1,2,…,n, are positive weights
that reflect the decision maker’s preferences
regarding the relative importance of each goal.
Example – 4
 TopAd, a new advertising agency with 10 employees, has
received a contract to promote a new product. The agency can
advertise by radio and television. The following table gives
the number of people reached by each type of advertisement
and the cost and labor requirements.
-------------------------------------------------------------------------------------------
Data/min advertisement
-------------------------------------
Radio Television
------------------------------------------------------------------------------------------
Exposure (in millions of persons) 4 8
Cost (in thousands of dollars) 8 24
Assigned employees 1 2
Example – 4 (Continued…)
 The contract prohibits TopAd from using more than 6
minutes of radio advertisement. Additionally, radio and
television advertisements need to reach at least 45 million
people. TopAd has a budget goal of $100,000 for the project.
How many minutes of radio and television advertisement
should TopAd use?
 Let x1 and x2 be the minutes allocated to radio and television
advertisements.
Example – 4 (Continued…)
 The goal programming formulation for the problem is given
as.
Minimize G1 = d1- (Satisfy exposure goal)
Minimize G2 = d2+ (Satisfy budget goal)
Subject to
4x1 + 8 x2 –d1+ + d1- = 45 (Exposure Goal)
8x1 + 24x2 –d2+ + d2- = 100 (Budget Goal)
x1+ 2x2 < 10 (Personnel limit)
x1 < 6 (Radio limit)
x1, x2, d1+ , d1- , d2+ , d2- > 0
Example – 4 (Continued…)
 TopAd’s management assumes that the exposure goal is twice
as important as the budget goal. The combined objective
function thus becomes:
Minimize z = 2G1 + G2 = 2d1- + d2+
 The preemptive method
 In the preemptive method, the decision maker must
rank the goals of the problem in order of
importance. Given an n-goal situation, the
objectives of the problem are written as
Minimize G1 = p1 (Highest priority)
.
.
Minimize Gn = pn (Lowest priority)

 The variable pi is the component of the deviational

variables, di+ or di- that represents goal i.
 In the Example-4 model , p1 d1- and p2 d2+
The preemptive method
 The solution procedure considers one goal at a time, starting
with the highest priority, G1, and terminating with the
lowest , Gn.
 The process is carried out such that the solution obtained
from a lower-priority goal never degrades any higher
priority solution.
 Previous "optimal" achievements of goals are added to the
constraint set so that they are not degraded while trying to
achieve lesser priority goals.
The preemptive method
 Step-0: Identify the goals of the model and rank them in
order of priority
G1=p1 > G2=p2 > … > Gn=pn

 Step-1:
o Solve LPi that minimizes Gi, and let pi = pi* define the
corresponding optimum value of the deviational variable pi.
o If i = n, stop; LPn solves the n-goal program.
o Otherwise, add the constraint pi = pi* to the constraints of the
Gi-problem to ensure that the value of pi will not be degraded
in future problem.
Set i = i+1, and repeat step i.
Example 5: Religare Securities Ltd
 Religare Securities Ltd has a client with $80,000 to invest
in a portfolio of only two stocks:

Stock Price/Share Estimated Risk

Annual Index/Share
Return/Share
United $25 $3 0.50
Breweries(U)
HUL (H) $50 $5 0.25

 The client has TWO objectives:

• Priority 1: Find a portfolio with a risk index of 700 or less
• Priority 2: Find a portfolio that will provide a return of $9000+
 Determine the goal programming formulation summary
Step 1: Identify any functional/resource constraints (not goal related)
Step 2: For each priority level, determine the goal and associated
constraints
Constraints will include the deviations di+ (amount greater than target)
and/or di- (amount less than target), from the goal
Each goal will minimize the undesired deviation.

Available funds constraint:

$25U + $50H < $80,000
Priority one level (portfolio risk of 700 or less):
Constraint: 0.50U + 0.25H = 700 - d1- + d1+
Goal: Min d1+
Priority two goal (annual return of $9000+):
Constraint: $3U + $5H = $9000 - d2- + d2+
Goal: Min d2-
 Step 3: Set up the formulation summary
• Objective Function: Minimize sum of all undesired deviations (denote
priority level of deviation with P1, P2, etc.; with 1 being highest level of
priority)
• Constraint Set consists of
• All resource constraints
• All goal-related constraints (all equalities with deviations)
• All non-negativity (all decision variables and deviations)

Min P1(d1+) + P2(d2-)

s.t., 25U + 50H < 80,000 Funds available

0.50U + 0.25H +d1- -d1+ = 700 P1 goal (Risk)

3U + 5H +d2- -d2+ = 9000 P2 goal (Return)

U, H, d1-, d1+, d2-, d2+ > 0

NOTE: Here, P1 and P2 are not weights, they are labels that indicate priority
Solution
Graph

Priority 1 Goal

Resource Constraint

Priority 2 Goal
Example – 4 (Revisited)
 The goal programming formulation for the problem is given
as.
Minimize G1 = d1- (Satisfy exposure goal)
Minimize G2 = d2+ (Satisfy budget goal)
Subject to
4x1 + 8 x2 + d1- –d1+ = 45 (Exposure Goal)
8x1 + 24x2 + d2- –d2+ = 100 (Budget Goal)
x1+ 2x2 < 10 (Personnel limit)
x1 < 6 (Radio limit)
x1, x2, d1+ , d1- , d2+ , d2- > 0
Example – 4 (Preemptive method)
 Assume that the exposure goal has a higher priority.
 Step 1. G1 > G2
G1 = Minimize d1- (Satisfy exposure goal)
G2 = Minimize d2+ (Satisfy budget goal)
 Step 2. Solve LP1
Minimize G1 = d1-
Subject to
4x1 + 8 x2 –d1+ + d1- = 45 (Exposure Goal)
8x1 + 24x2 –d2+ + d2- = 100 (Budget Goal)
x1+ 2x2 < 10 (Personnel limit)
x1 < 6 (Radio limit)
x1, x2, d1+ , d1- , d2+ , d2- > 0
Example – 4 (Preemptive method)
 The optimal solution is x1=5, x2=2.5 and d1- = 5 with remaining
variables equal to zero.
 Step 3. Solve LP2
Minimize G2 = d2+
Subject to
4x1 + 8 x2 –d1+ = 40 (Exposure Goal) , (d1- = 5 )
8x1 + 24x2 –d2+ + d2- = 100 (Budget Goal)
x1+ 2x2 < 10 (Personnel limit)
x1 < 6 (Radio limit)
x1, x2, d1+ , d2+ , d2- > 0

 The optimization of LP2 is not necessary, because d2+ is already

zero (minimum) in LP1.
 Hence LP1 solution is optimal.
 Graphical Solution: Summary
 Step 1. Graph or plot all structural constraints and identify the
feasible region. If no structural constraints exist, the nonnegative
quadrant is the feasible region.
 Step 2. Graph the lines corresponding to the goal constraints by
setting deviational variables to zero.
 Step 3. Identify the top priority solution. This is accomplished by
determining the point or points within the feasible region that
satisfy the highest priority goal.
 Step 4. Move to the goal having next highest priority and
determine the best solution (s), such that this ‘best’ solution does
not degrade the solution(s) already achieved for goals of higher
priority.
 Step 5. Repeat step 4 until all priority levels have been achieved.
Example 6: Conceptual Products
Conceptual Products is a computer company that
produces the CP400 and CP500 computers. Many of
the components used in the two
computer models are produced in
abundant supply by the company.
However, the memory modules,
external hard drives, and cases are
bought from suppliers.
The CP400 model uses two memory modules and
no external hard drive, whereas the CP500 uses one
memory module and one external hard drive. Both
models use one case.
Example 6: Conceptual Products
Suppliers can provide Conceptual Products with 1000 memory
modules, 500 external hard drives, and 600 cases on a weekly basis.
It takes one hour to manufacture a CP400 and its profit is $200 and
it takes one and one-half hours to manufacture a CP500 and its
profit is $500.

The company has four goals:

Priority 1: Meet a contract of 200 CP400 machines weekly. (Goal 1)

Priority 2: Make at least 500 total computers weekly. (Goal 2)
Priority 3: Make at least $250,000 weekly. (Goal 3)
Priority 4: Use no more than 400 man-hours per week. (Goal 4)
EX6. Goal Programming: Formulation
 Variables
x1 = number of CP400 computers produced weekly
x2 = number of CP500 computers produced weekly
di- = amount the right hand side of goal i is deficient
di+ = amount the right hand side of goal i is exceeded

 Functional Constraints
Availability of memory modules: 2x1 + x2 < 1000
Availability of external hard drives: x2 < 500
Availability of cases: x1 + x2 < 600
EX6. Goal Programming: Formulation

 Goals
(1) 200 CP400 computers weekly:
x1 + d1- - d1+ = 200
(2) 500 total computers weekly:
x1 + x2 + d2- - d2+ = 500
(3) $250(in thousands) profit:
.2x1 + .5x2 + d3- - d3+ = 250
(4) 400 total man-hours weekly:
x1 + 1.5x2 + d4- - d4+ = 400
Non-negativity:
x1, x2, di-, di+ > 0 for all i
 EX6. Goal Programming: Formulation
 Objective Functions

Priority 1: Minimize the amount the state contract is not met:

Min d1-

Priority 2: Minimize the number under 500 computers produced

weekly: Min d2-

Priority 3: Minimize the amount under $250,000 earned weekly:

Min d3-

Priority 4: Minimize the man-hours over 400 used weekly:

Min d4+
EX6. Goal Programming: Formulation
 Formulation Summary

Min P1(d1-) + P2(d2-) + P3(d3-) + P4(d4+)

s.t. 2x1 +x2 < 1000

+x2 < 500
x1 +x2 < 600
x1 +d1- -d1+ = 200
x1 +x2 +d2- -d2+ = 500
.2x1+ .5x2 +d3- -d3+ = 250
x1+1.5x2 +d4- -d4+ = 400

x1, x2, d1-, d1+, d2-, d2+, d3-, d3+, d4-, d4+ > 0
 EX6. Goal Programming: Solution

Graphical Solution
 Iteration 1
To solve graphically, first graph the functional constraints.
Then graph the first goal: x1 = 200.
EX6. Goal Programming: Solution
 Functional Constraints and Goal 1 Graphed
x2
1000 2x1 + x2 < 1000

800 Goal 1: x1 > 200

x2 < 500
600 x1 + x2 < 600

400
Points
Satisfying
200
Goal 1
x1
200 400 600 800 1000 1200
 EX6. Goal Programming: Solution
 Iteration 2
Now keeping Goal 1 as x1 > 200 and graph Goal 2:
x1 + x2 = 500.
Note on the next slide that there is still a set of points satisfying the
first goal that also satisfies this second goal (i.e.; d2- = 0).
EX6. Goal Programming: Solution
 Goal 1 (Constraint) and Goal 2 Graphed
x2
1000
2x1 + x2 < 1000
800 Goal 1: x1 > 200
x2 < 500
600 x1 + x2 < 600

400 Points Satisfying

Both Goals 1 and 2
200 Goal 2: x1 + x2 > 500

x1
200 400 600 800 1000 1200
 EX6. Goal Programming: Solution
 Iteration 3
Now keeping Goal 1, Goal 2, plot Goal 3:
.2x1 + .5x2 = 250.
Note on the next slide that no points satisfy the previous functional
constraints and goals and satisfy this constraint.

Thus, to Min d3-, this minimum value is achieved when we Max .2x1 +
.5x2.
Note that this occurs at x1 = 200 and x2 = 400, so that .2x1 + .5x2 = 240
or d3- = 10.
 EX6. Goal Programming: Solution
 Goal 2 (Constraint), Goal 3 and Goal 4 Graphed
x2
2x1 + x2 < 1000
1000
Goal 1: x1 > 200
800
x1 + x2 < 600 x2 < 500
600
(200,400)
400
Goal 2: x1 + x2 > 500
200
Goal 3: .2x1 + .5x2 > 250
x1
200 400 600 800 1000 1200
Goal 4: x1 + 1.5x2 < 400
Example – 3 Revisited (An electric company):
(Preemptive method)
 The goal programming problem can be formulated as:

Min Z = p1d1- + p2d2- +p3d3+ +p4d4-

St
7x1+ 6x2 –d1+ + d1- = 30
2x1+ 3x2 –d2+ + d2- =12
6x1+ 5x2 –d3+ + d3- =30
x2 –d4+ + d4- = 7
x1, x2 ,di+, di- >0  i =1,2,3,4.
Example – 3(Continued…)
Example – 3(Continued…)
Example – 3(Continued…)
X1 = 0
X2 = 6
d1- = 0
d2- = 0
d3+ = 0
d4- = 1
Example 6: Burnit Goal Programming
 The Leon Burnit Adveritsing Agency is trying to determine a
TV advertising schedule for Priceler Auto Company.
 Priceler has three goals:
 Its ads should be seen by at least 40 million high-income
men (HIM).
 Its ads should be seen by at least 60 million low-income
people (LIP).
 Its ads should be seen by at least 35 million high-income
women (HIW).
 Leon Burnit can purchase two types of ads: those shown
during football games and those shown during soap operas.
Ex.6 - continued

 At most, $600,000 can be spent on ads.

 The advertising costs and potential audiences of a one-minute
ad of each type are shown.

Millions of Viewers
Ad HIM LIP HIW Cost (S)
Football 7 10 5 100,000
Soap Opera 3 5 4 60,000

 Leon Burnit must determine how many football ads and soap
opera ads to purchase for Priceler.
 Summary
 If a decision maker has a linear cost function, goal programming can be
used in two different ways to represent tradeoffs between different goals. In
both the methods, deviational variables are used to convert each goal into a
constraint.
 Assign a penalty per unit of deviation from each goal and use linear programming to
minimize the total penalty incurred because of unmet goals.
 Ranks goals in priority from highest to lowest and use preemptive goal programming and
the goal programming simplex.
 (Preemptive goal programming finds all points that come as close as
possible to meeting goal 1. Then among these points, find the points that
come as close as possible to meeting the next goal. Continue this until any
improvement in a goal can come only at the expense of a higher-priority
goal.)
Multi Objective Optimization
There exists many methods to solve MOO problems.
 For example, Evolutionary Algorithms using Genetic Algorithms, Neural
Computing, etc.
 Evolutionary algorithms seem particularly suitable to solve multi-objective optimization
problems, because they deal simultaneously with a set of possible solutions (the so-called
population). This allows us to find several members of the Pareto optimal set in a single
run of the algorithm, instead of having to perform a series of separate runs as in the case
of the traditional mathematical programming techniques.
 Simulated Annealing
 Epsilon Constraint Methods
 Posterior Preference
 Particle Swarm Optimization
 …