You are on page 1of 60

Chapter 3

Multi Criteria Decision Making

Yitbarek Takele (PhD, MBA & MA Econ)


Associate Professor of Management
Addis Ababa University
3.1 Introduction
 A company or an organization often has more
than one objective, which may relate to
something other than profit or cost
 In fact, a company may have several
criteria, that is, multiple criteria, that it will
consider in making a decision instead of just a
single objective.

2
• In addition to maximizing profit, a company in
danger of a labor strike might want to avoid
employee layoffs, or a company about to be
fined for pollution infractions might want to
minimize the emission of pollutants
• In this chapter we discuss three techniques
that can be used to solve problems when they
have multiple objectives: goal programming,
the analytical hierarchy process, and
scoring models.

3
3.2 Goal Programming
• Goal programming models are set up in the same
general format as linear programming models,
with an objective function and linear
constraints
• Goal programming model is very similar to a
linear programming model, with an objective
function, decision variables, and constraints.
• Like linear programming, goal programming
models with two decision variables can be solved
graphically and by using LINDO/LINGO

4
Model Formulation
 Beaver Creek Pottery Company is a small crafts
operation run by a Native American tribal council.
The company employs skilled artisans to produce
clay bowls and mugs with authentic Native
American designs and colors. The two primary
resources used by the company are special
pottery clay and skilled labor. Given these
limited resources, the company desires to know
how many bowls and mugs to produce each day in
order to maximize profit. This is generally referred
to as a product mix problem type.

5
 Max z=40x1 +50x2
 Subject to:
x1+2x2<40hr of labor
4x1+3x2<120lb of clay
x1,x2>0

 Where
X1=number of bowls produced
x2= number of mugs produced

6
 This is a standard linear programming model; as
such, it has a single objective function for profit.
However, let us suppose that instead of having one
objective, the pottery company has several
objectives, listed here in order of importance:
 To avoid layoffs, the company does not want to use
fewer than 40 hours of labor per day.
 The company would like to achieve a satisfactory profit
level of $1,600 per day.
 Because the clay must be stored in a special place so
that it does not dry out, the company prefers not to
keep more than 120 pounds on clay each day.
 Because high overhead costs result when the plant is
kept open past normal hours, the company would like to
minimize the amount of over time. 7
 The different objectives in a goal programming
problem are referred to as goals
 Our first step in formulating a goal programming
model is to transform the linear programming model
constraint into goals.
 Labor Goal- The first goal of the pottery company
is to avoid underutilization of labor that is, using
fewer than 40 hours of labor each day.
 To represent the possibility of underutilizing labor,
the linear programming constraint for labor, x1 +
2x2 <40 hours of labor, is reformulated as
 X1+2x2+d1- - d1+ = 40hr
8
 This reformulated equation is referred to as a
goal constraint. The two new variables, d1- and
d1+, are called deviational variables. They
represent the number of labor hours less than 40
(d1-) and the number of labor hours exceeding 40
(d1+). More specifically, d1- represents labor
underutilization, and d1+ represents overtime

9
 All goal constraints are equalities that include
deviational variables, d- and d+.
 A positive deviational variable (d+) is the amount
by which a goal level is exceeded.
 A negative deviational variable (d-) is the amount
by which a goal level is underachieved
 At least one or both deviational variables in a goal
constraint must equal zero

10
 The next step in formulating our goal programming
model is to represent the goal of not using fewer
than 40 hours of labor. We do this by creating a
new form of objective function:
 Labor Goal: P1d1-, P4d1+

11
 Min P 1d 1-
 The objective function in all goal programming models
is to minimize deviation from the goal constraint
levels. In this objective function, the goal is to
minimize d1-, the underutilization of labor. If d1-
equaled zero, then we would not be using fewer than
40 hours of labor. Thus, it is our objective to make d1-
equal zero or the minimum amount possible. The
symbol P1 in the objective function designates the
minimization of d1- as the first-priority goal. This
means that when this model is solved, the first step
will be to minimize the value of d1- before any
other goal is addressed.
12
 The objective function in a goal programming
model seeks to minimize the deviation from goals
in order of the goal priority
 Min P4d1+
 The objective is to minimize the deviational variable
d1+. In other words, if d1+ equaled zero, there
would be no overtime at all. In solving this model,
the achievement of this fourth-ranked goal will not
be attempted until goals one, two, and three have
been considered

13
 Profit Goal: P2d2-
 The second goal in our goal programming model is to
achieve a daily profit of $1,600. Recall that the original
linear programming objective function was
Z = 40x1 + 50x2
 Now we reformulate this objective function as a goal
constraint with the following goal level:
40x1 + 50x2 + d2- - d2+ = $1,600
 The deviational variables d2- and d2+ represent the amount
of profit less than $1,600 (d2-) and the amount of profit
exceeding $1,600 (d2+). The pottery company's goal of
achieving $1,600 in profit is represented in the objective
function as minimize p1d1-, p2d2-,p4d1+
14
 Notice that only d2- is being minimized, not d2+,
because it is logical to assume that the pottery
company would be willing to accept all profits in
excess of $1,600 (i.e., it does not desire to
minimize d2+, excess profit). By minimizing d2- at
the second-priority level, the pottery company
hopes that d2- will equal zero, which will result in at
least $1,600 in profit

15
Material Goal: P3d3+
 Goal 3:
 The third goal of the company is to avoid keeping more
than 120 pounds of clay on hand each day. The goal
constraint is
4x1 + 3x2 + d3- - d3+ = 120 lb.
 Because the deviational variable d3- represents the amount
of clay less than 120 pounds, and d3+ represents the
amount in excess of 120 pounds, this goal can be reflected
in the objective function as:
 Overall goal (G1, G2, G3)
 Min p1d1-, p2d2-, p3d3+, p4 d1+
16
 The term P3d3+ represents the company's desire to
minimize d3+, the amount of clay in excess of 120
pounds. The P3 designation indicates that it is the
pottery company's third most important goal.
 The complete goal programming model can now be
summarized as follows:

17
 Min P1d1-, P2d2-, P3d3+ , P4d1+
Subject to:
X1+2x2+d1- - d1+ =40
40x1+50x2+d2- - d2+ =1600
4x1+3x2+d3- - d3+= 120
X1,X2, d1- , d1+ +, d2- , d2+ , d3- , d3+ >0

18
 The one basic difference between this model and
the standard linear programming model is that the
objective function terms are not summed to
equal a total value, Z. This is because the
deviational variables in the objective function
represent different units of measure
 For example, d1- and d1+ represent hours of labor, d2-
represents dollars, and d3+ represents pounds of clay. It
would be illogical to sum hours, dollars, and pounds. The
objective function in a goal programming model specifies
only that the deviations from the goals represented in the
objective function be minimized individually, in order
of their priority.

19
3.3 The Analytical Hierarchy Process
(AHP)
 Goal programming is a method that provides us
with a mathematical "quantity" for the
decision variables that best achieves a set of
goals.
 It answers the question "How much?"
 The analytical hierarchy process (AHP), developed
by Thomas Saaty, is a method for ranking
decision alternatives and selecting the best
one when the decision maker has multiple
objectives, or criteria, on which to base the
decision.

20
 Thus, it answers the question "Which one?"
 A decision maker usually has several alternatives
from which to choose when making a decision
 For example, someone buying a house might have
several houses for sale from which to choose;
 someone buying a new car might have several makes and
styles to consider; and
 a prospective student might select a college to attend from
a group of schools
 In each of these examples, the decision maker
would typically make a decision based on how the
alternatives compare, according to several
21
criteria.
 AHP is a process for developing a numeric score
to rank each decision alternative, based on
how well each alternative meets the decision
maker's criteria
 E.g. Southcorp Development builds and operates shopping
malls in America. The company has identified three
potential sites for its latest project, near Atlanta,
Birmingham, and Charlotte. The company has identified
four primary criteria on which it will compare the sites:
(1) the customer market base (including overall market
size and population at different age levels); (2) income
level; (3) infrastructure (including utilities and roads); and
(4) transportation (i.e., proximity to interstate highways
for supplier deliveries and customer access).
22
Hierarchy Levels
 1st level of the hierarchy
 Set the overall objective: select the best site.
 This goal is at the top of the hierarchy of the problem.
 2nd level of the hierarchy:
 Determine how the four criteria contribute to
achieving the objective.
 3rd level of the hierarchy:
 Determine how each of the three alternatives (Atlanta,
Birmingham, and Charlotte) contributes to each of the
four criteria.
23
 We mathematically determine our preferences for
each site for each criterion.
 General steps:
i. Determine how well each alternative “Scores" on a criterion
by using pair-wise comparisons
ii. Developing Preferences within Criteria
iii. Ranking the criteria
iv. Developing an Overall Ranking

24
Step-1: Pair-wise Comparison b/n
Alternatives
 In a pair-wise comparison, the decision maker
compares two alternatives (i.e., a pair) according to
one criterion and indicates a preference. For
example, Southcorp might compare the Atlanta (A) site
with the Birmingham (B) site and decide which one it
prefers according to the customer market criterion.
These comparisons are made by using a preference
scale, which assigns numeric values to different levels
of preference.
 In a pair-wise comparison, two alternatives are
compared according to a criterion, and one is preferred
 The standard preference scale used for AHP is shown
with numeric scale 1 to 9
25
Preference Scale for Pair-wise Comparisons

Preference Level Numeric Value


 Equally preferred 1
 Equally to moderately preferred 2
 Moderately preferred 3
 Moderately to strongly preferred 4
 Strongly preferred 5
 Strongly to very strongly preferred 6
 Very strongly preferred 7
 Very strongly to extremely preferred 8
 Extremely preferred 9

26
 If Southcorp compares Atlanta to Birmingham
and moderately prefers Atlanta, resulting in a
comparison value of 3 for the customer market
criterion, it is not necessary for Southcorp to
compare Birmingham to Atlanta to determine a
separate preference value for this "opposite"
comparison. The preference value of B for A is
simply the reciprocal or inverse of the preference of
A for B. Thus, in our example if the preference
value of Atlanta for Birmingham is 3, the
preference value of Birmingham for Atlanta is 1/3.

27
 Southcorp's pair-wise comparison ratings for each
of the three sites for the customer market criterion
are summarized in a matrix, a rectangular array of
numbers. This pair-wise comparison matrix will
have a number of rows and columns equal to the
decision alternatives:

28
Site/Customer A B C
Market Size
A 1 3 2
B 1/3 1 1/5
C 1/2 5 1

A pair-wise comparison matrix summarizes the


pair-wise comparisons for a criterion.

29
 This matrix shows that the customer market in
Atlanta (A) is equally to moderately preferred (2)
over the Charlotte (C) customer market, but
Charlotte (C) is strongly preferred (5) over
Birmingham (B). Notice that any site compared
against itself, such as A compared to A, must be
"equally preferred," with a preference value of 1.
Thus, the values along the diagonal of our matrix
must be 1s

30
 The remaining pair-wise comparison matrices for
the other three criteria income level,
infrastructure, and transportation have been
developed by Southcorp as follows

Income Level Infrastructure Transportation


A B C A B C A B C
A 1 6 1/3 1 1/3 1 1 1/3 1/2
B 1/6 1 1/9 3 1 7 3 1 4
C 3 9 1 1 1/7 1 2 ¼ 1

31
Step-2:Developing Preferences within
Criteria
 The next step in AHP is to prioritize the decision
alternatives within each criterion. For our site
selection example, this means that we want to
determine which site is the most preferred, the
second most preferred, and the third most
preferred for each of the four criteria. This step in
AHP is referred to as synthesization.
 The mathematical procedure for same is called
synthesization
 In synthesization, decision alternatives are
prioritized within each criterion

32
 We will use an approximation method for
synthesization that provides a reasonably good
estimate of preference scores for each decision in
each criterion.
Steps:
1. Sum the values in each column of the pair-wise
comparison matrices. The column sums for our
customer market matrix are shown as follows:

33
Customer Market Size
Site A B C
A 1 3 2
B 1/3 1 1/5
C ½ 5 1
11/6 9 16/5
2. Divide each value in a column by its corresponding
column sum.
• The results in a normalized matrix are shown as
follows:
34
Customer Market Size
Site A B C
A 6/11 3/9 5/8
B 2/11 1/9 1/16
C 3/11 5/9 5/16
 Notice that the values in each column sum to 1.
3. Average the values in each row. At this point,
we convert the fractional values in the matrix to
decimals, as shown next.

35
Customer Market Size
Site A B C Row averages

A 0.5455 0.3333 0.6250 0.5012


B 0.1818 0.1111 0.0625 0.1185
C 0.2727 0.5556 0.3125 0.3803
1.0000
 The row averages shown above provide us with
Southcorp's preferences for the three sites for the
customer market criterion. The most preferred site is
Atlanta, followed by Charlotte; the least preferred site
(for this criterion) is Birmingham. We can write these
preferences as a matrix with one column, which we will
refer to as a vector: 36
 These four preference vectors for the four criteria
are summarized in a single preference matrix

Criteria Preference Matrix


Criterion

Site Market Income level Infrastructure Transportation


A 0.5012 0.2819 0.1790 0.1561
B 0.1185 0.0598 0.6850 0.6196
C 0.3803 0.6583 0.1360 0.2243

37
Step-3: Ranking Criteria
 The next step in AHP is to determine the relative
importance, or weight, of the criteria that is, to
rank the criteria from most important to least
important. This is accomplished the same way we
ranked the sites within each criterion: using
pairwise comparisons. The following pairwise
comparison matrix for the four criteria in our
example was developed by using the preference
scale

38
Criterion
Market Income Infrastructure Transportation
Market 1 1/5 3 4
Income 5 1 9 7
Infrastructure 1/3 1/9 1 2
Transportation 1/4 1/7 1/2 1

 The normalized matrix converted to decimals, with the


row averages for each criterion

39
Criterion: Market Income Infra Transp Row Av.
Market 0.1519 0.1375 0.2222 0.2857 0.1993
Income 0.7595 0.6878 0.6667 0.5000 0.6535
Infrra 0.0506 0.0764 0.0741 0.1429 0.0860
Transp 0.0380 0.0983 0.0370 0.0714 0.0612
1.0000
The preference vector, computed from the normalized matrix by
computing the row averages
Criterion
Market 0.1993
Income 0.6535
Infrastructure 0.0860
Transportation 0.0612
40
 Clearly, income level is the highest-priority
criterion, and customer market is second.
Infrastructure and transportation appear to be
relatively unimportant third- and fourth-ranked
priorities in terms of the overall objective of
determining the best site for the new shopping mall
 The next step in AHP is to combine the preference
matrices we developed for the sites for each
criterion with the preceding preference vector for
the four criteria

41
Site Criterion

Market Income Infrast. Transp. Market 0.1993


A 0.5012 0.2819 0.1790 0.1561 Income 0.6535
B 0.1185 0.0598 0.6850 0.6196 Infrastructure 0.0860
C 0.3803 0.6583 0.1360 0.2243 Transportation 0.0612

42
Step-4: Develop an Overall Ranking
 An overall score for each site is computed by
multiplying the values in the criteria preference
vector by the preceding criteria matrix and
summing the products, as follows:
 Site A score = 0.1993(0.5012) + 0.6535(0.2819) + 0.0860(0.1790) +
0.0612(0.1561) = 0.3091
 Site B score = 0.1993(0.1185) + 0.6535(0.0598) + 0.0860(0.6850) +
0.0612(0.6196) = 0.1595
 Site C score = 0.1993(0.3803) + 0.6535(0.6583) + 0.0860(0.1360) +
0.0612(0.2243) = 0.5314

43
 The three sites, in order of the magnitude of their
scores, result in the following AHP ranking:
Site Score
 Charlotte 0.5314
 Atlanta 0.3091
 Birmingham 0.1595
1.0000
 Based on these scores developed by AHP, Charlotte
should be selected as the site for the new
shopping mall, with Atlanta ranked second and
Birmingham third.
44
 Following is a summary of the mathematical steps
used to arrive at the AHP-recommended decision:
1.Develop a pair-wise comparison matrix for each
decision alternative (site) for each criterion.
2. Synthesization:
Sum the values in each column of the pair-wise
comparison matrices.
 Divide each value in each column of the pair-wise comparison
matrices by the corresponding column sum these are the
normalized matrices.
 Average the values in each row of the normalized matrices
these are the preference vectors.
 Combine the vectors of preferences for each criterion (from
step 2) into one preference matrix that shows the
preference for each site for each criterion. 45
3. Develop a pair-wise comparison matrix for the
criteria.
4. Compute the normalized matrix by dividing each value
in each column of the matrix by the corresponding column
sum.
5. Develop the preference vector by computing the row
averages for the normalized matrix.
6. Compute an overall score for each decision alternative by
multiplying the criteria preference vector (from step
5) by the criteria matrix (from step 2d).
7. Rank the decision alternatives, based on the magnitude
of their scores computed in step 6

46
Checking for Consistency
 AHP is based primarily on the pair-wise
comparisons a decision maker uses to establish
preferences between decision alternatives for
different criteria. The normal procedure in AHP
for developing these pair-wise comparisons is for
an interviewer to elicit verbal preferences from the
decision maker, using the preference scale.
 However, when a decision maker has to make a lot
of comparisons (i.e., three or more), he or she can
lose track of previous responses. Because AHP is
based on these responses, it is important that
they be in some sense valid and especially
that the responses be consistent.
47
 A consistency index measures the degree of
inconsistency in pair-wise comparisons.
 To demonstrate how to compute the consistency
index (CI), we will check the consistency of the
pair-wise comparisons for the four site selection
criteria. This matrix, shown as follows, is
multiplied by the preference vector for the
criteria:

48
Site Market Inc level Infr. transp criteria
Market 1 1/5 3 4 0.1993 0.8328
Inc.levl. 5 1 9 7 0.6535 = 2.8524
Inf. 1/3 1/9 1 2 * 0.0860 0.3474
Trans. ¼ 1/7 ½ 1 0.0612 0.2473

49
 Next, we divide each of these values by the
corresponding weights from the criteria preference
vector:
0.8328/0.1993= 4.1786
2.8524/0.6535= 4.3648
0.3474/0.0860= 4.0401
0.2473/0.0612= 4.0422
16.257

 If the decision maker, Southcorp, were a perfectly


consistent decision maker, then each of these ratios
would be exactly 4, the number of items we are
comparing in this case, four criteria.
50
 Next, we average these values by summing them
and dividing by 4:
16.6257/4=4.1564
 The consistency index, CI, is computed using the
following formula:
 CI= 4.1564 - n
n-1
where
n = the number of items being compared
4.1564 = the average we computed previously
CI= 4.1564-4 = 0.0521
3
51
 If CI = 0, then Southcorp would be a perfectly
consistent decision maker. Because Southcorp is
not perfectly consistent, the next question is the
degree of inconsistency that is acceptable.
 An acceptable level of consistency is determined by
comparing the CI to a random index, RI, which is
the consistency index of a randomly generated
pairwise comparison matrix
 The RI has the values depending on the number of
items, n, being compared. In our example, n = 4
because we are comparing four criteria.

52
N 2 3 4 5 6 7 8 9 10
RI 0 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.5
 The degree of consistency for the pair-wise
comparisons in the decision criteria matrix is
determined by computing the ratio of CI to RI:
CI/RI=0.0521/0.90=0.0580
 In general, the degree of consistency is satisfactory
if CI/RI < 0.10, and in this case, it is. If CI/RI >
0.10, then there are probably serious
inconsistencies, and the AHP results may not be
meaningful.
53
3.4 Scoring Models
 For selecting among several alternatives according
to various criteria, a scoring model is a method
similar to AHP, but it is mathematically simpler.
There are several versions of scoring models. In the
scoring model that we will use, the decision
criteria are weighted in terms of their relative
importance, and each decision alternative is
graded in terms of how well it satisfies the
criteria, according to the following formula:

54
 Si = ∑gij wj
Where:
 Wj=a weight between 0 and 1.00 assigned to criterion
j, indicating its relative importance, where 1.00 is
extremely important and 0 is not important at all. The
sum of the total weights equals 1.00.
 Gij=a grade between 0 and 100 indicating how well
the decision alternative i satisfies criterion j,
where 100 indicates extremely high satisfaction and 0
indicates virtually no satisfaction.
 Si=the total "score" for decision alternative i, where
the higher the score, the better.
55
 Example
 Sweats and Sweaters is a chain of stores specializing in
cotton apparel. The company wants to open a new store in
one of four malls around the Atlanta metropolitan area.
The company has indicated five criteria that are important
in its decision about where to locate: proximity of schools
and colleges, area median income, mall vehicle traffic flow
and parking, quality and size (in terms of number of stores
in the mall), and proximity of other malls or shopping
areas. The company has weighted each of these criteria in
terms of its relative importance in the decision-making
process, and it has analyzed each potential mall location
and graded them according to each criterion as shown in
the following table:

56
 Grades for Alternative (0 to 100)
 Decision Criteria Weight (0 to 1.00)
Weight Mall 1 Mall 2 Mall 3 Mall 4
School proximity 0.30 40 60 90 60
Median income 0.25 75 80 65 90
Vehicle traffic 0.25 60 90 79 85
Mall quality and size 0.10 90 100 80 90
Prox. of other shopping 0.10 80 30 50 70

 The scores, Si, for the decision alternative are computed as follows:

57
S1 = (.30)(40) + (.25)(75) + (.25)(60) + (.10)(90) + (.10)(80) = 62.75
S2 = (.30)(60) + (.25)(80) + (.25)(90) + (.10)(100) + (.10)(30) = 73.50
S3 = (.30)(90) + (.25)(65) + (.25)(79) + (.10)(80) + (.10)(50) = 76.00
S4 = (.30)(60) + (.25)(90) + (.25)(85) + (.10)(90) + (.10)(70) = 77.75

 Because Mall 4 has the highest score, it would be


the recommended decision, followed by Mall 3,
Mall 2, and finally Mall 1.

58
Discussion

Session
59
I Thank You

60

You might also like