Analytic Hierarchy Process

Multi-criteria decision making.

 The Analytic Hierarchy Process (AHP)

It is popular and widely used method for

multi-criteria decision making.
Allows the use of qualitative, as well as
quantitative criteria in evaluation.
Founded by Saaty in 1980.

2
Typical application areas
Resource allocation
Hiring, evaluating and promoting employees
TQM
Strategic planning
Relocation decisions
Vendor selection
Evaluating mergers and acquisitions
Selecting a car for purchasing
Deciding upon a place to visit for vacation
Deciding upon an MBA program after graduation.
 How widespread is its use?
IBM NASA
Goodyear IRS
Ford Motor Co. FBI
..a few of the thousands of
Citibank Department of
organizations using AHP Defense
and EC
Xerox World Bank
Boeing Texaco
AT&T Eastman Kodak
General Motors Inter-American
Bank
 AHP-General Idea
Develop an hierarchy of decision criteria and define the
alternative courses of actions.

AHP algorithm is basically composed of two steps:

1. Determine the relative weights of the decision criteria
2. Determine the relative rankings (priority) of alternatives

! Both qualitative and quantitative information can be

compared using informed judgements to derive weights
and priorities.

5
Example: Car Selection
Objective
Selecting a car
Criteria
Style, Reliability, Fuel-economy Cost?
Alternatives
Civic Coupe, Saturn Coupe, Ford Escort,
Mazda Miata

6
Analytic Hierarchy Process
Step 1: Structure a hierarchy. Define the problem,
determine the criteria and identify the alternatives.
Overall Goal Select the Best Toothbrush Manufacturer

Criteria Cost Reliability Delivery Time

Cornell Cornell Cornell

Decision Brush Pik Brush Pik
Brush Pik
Alternatives Picobuy Picobuy
Picobuy
 Hierarchy tree

Selecting
a New Car

Style Reliability Fuel Economy

Civic Saturn Escort Miata

Alternative courses of action

8
Analytic Hierarchy Process

Step 2: Make pairwise comparisons. Rate

the relative importance between each pair
of decision alternatives and criteria.
 Analytic Hierarchy Process
Step 2 (contd): AHP uses 1-9 scale for the
prioritization process.
Numerical ratings Verbal judgments
1 Equally important (preferred)
3 Moderately more important
5 Strongly more important
7 Very strongly more important
9 Extremely more important
11
Analytic Hierarchy Process
Step 2 (contd): Intermediate numerical
ratings of 2, 4, 6, and 8 can be assigned. If
someone could not decide whether one
criterion (alternative) is moderately more
important than the other one or strongly
more important than the other one, 4
(moderately to strongly more important)
can be assigned.
Analytic Hierarchy Process

Step 3: Synthesize the results to determine

the best alternative. Obtain the final results.
The output of AHP is the set of priorities of
the alternatives.
An Example with AHP
Choosing the most satisfied school
Goal: To select new car.
Criteria: price, MPG, comfort and style
Alternatives: car A, car B, car C
 Hierarchy Development
The first step in the AHP is to develop a graphical
representation of the problem in terms of the overall
goal, the criteria, and the decision alternatives.

Overall Goal: Select the Best Car

Criteria: Price MPG Comfort Style

Decision Car A Car A Car A Car A

Alternatives:
Car B Car B Car B Car B
Car C Car C Car C Car C
 Example: Synthesizing Procedure - 1
Step 1: Sum the values in each column.

Comfort Car A Car B Car C

Car A 1 2 8
Car B 1/2 1 6
Car C 1/8 1/6 1
Column totals 13/8 19/6 15
 Example: Synthesizing Procedure - 2
Step 2: Divide each element of the matrix by its column
total.
All columns in the normalized pairwise comparison
matrix now have a sum of 1.

Comfort Car A Car B Car C

Car A 8/13 12/19 8/15
Car B 4/13 6/19 6/15
Car C 1/13 1/19 1/15
 Example: Synthesizing Procedure - 3
Step 3: Average the elements in each row.
The values in the normalized pairwise comparison
matrix have been converted to decimal form.
The result is usually represented as the (relative)
priority vector (eigen vector).
Comfort Car A Car B Car C Row Avg.

Car A 0.615 0.632 0.533 0.593 0.593

0.341
Car B 0.308 0.316 0.400 0.341
0.066
Car C 0.077 0.053 0.067 0.066
Total 1.000
 Ranking of priorities
Consider [Ax = maxx] where
A is the comparison matrix of size nn, for n criteria, also called the priority matrix.
x is the Eigenvector of size n1, also called the priority vector.
max is the Eigenvalue, max > n.

To find the ranking of priorities, namely the Eigen Vector X:

1) Normalize the column entries by dividing each entry by the sum of the column.
2) Take the overall row averages.

Row
Normalized 0.30 0.29 0.38 0.30
1 0.5 3 Column Sums
averages
A= 2 1 4 0.60 0.57 0.50 0.60
X=
0.33 0.25 1.0 0.10 0.14 0.13 0.10
Priority vector
Column sums 3.33 1.75 8.00 1.00 1.00 1.00

Eigen Vector

20
Consistency - 1
An important consideration in terms of the
quality of the ultimate decision relates to the
consistency of judgments that the decision
maker demonstrated during the series of
pairwise comparisons.
It should be realized perfect consistency is very
difficult to achieve and that some lack of consistency
is expected to exist in almost any set of pairwise
comparisons.
Example:
Consistency - 2
To handle the consistency question, the AHP
provides a method for measuring the degree of
consistency among the pairwise judgments
provided by the decision maker.
If the degree of consistency is acceptable, the decision
process can continue.
If the degree of consistency is unacceptable, the
decision maker should reconsider and possibly revise
the pairwise comparison judgments before
proceeding with the analysis.
Consistency Ratio
The AHP provides a measure of the
consistency of pairwise comparison
judgments by computing a consistency
ratio
The ratio is designed in such a way that values
of the ratio exceeding 0.10 are indicative of
inconsistent judgments.
Although the exact mathematical
computation of the consistency ratio is
beyond the scope of this text, an
approximation of the ratio can be obtained.
Procedure: Estimating Consistency Ratio - 1

Step 1: Multiply each value in the first column of

the pairwise comparison matrix by the
relative priority of the first item considered.
Same procedures for other items. Sum the
values across the rows to obtain a vector
of values labeled weighted sum.
Step 2: Divide the elements of the vector of
weighted sums obtained in Step 1 by the
corresponding priority value.
Step 3: Compute the average of the values
computed in step 2. This average is
denoted as max.
Procedure: Estimating Consistency Ratio - 2
Step 4: Compute the consistency index (CI):
max n
CI
n 1
Where n is the number of items being compared

Step 5: Compute the consistency ratio (CR):

CI
CR
RI
Where RI is the random index, which is the consistency
index of a randomly generated pairwise comparison
matrix. It can be shown that RI depends on the number
of elements being compared and takes on the following
values.
Example:
 Random Index
Random index (RI) is the consistency index of a
randomly generated pairwise comparison matrix.
RI depends on the number of elements being
compared (i.e., size of pairwise comparison matrix) and
takes on the following values:

n 1 2 3 4 5 6 7 8 9 10

RI 0.00 0.00 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49
 Example: Inconsistency
Preferences: If, A B (2); B C (6)
Then, A C (4) (should be 8)

Inconsistency

Comfort Car A Car B Car C

Car A 1 2 8
Car B 1/2 1 6
Car C 1/8 1/6 1
 Example: Consistency Checking - 1
Step 1: Multiply each value in the first column of the pairwise
comparison matrix by the relative priority of the first
item considered. Same procedures for other items. Sum
the values across the rows to obtain a vector of values
labeled weighted sum.

1 2 8 0.593 0.682 0.528 1.803

0.593 1 2 0.341 1 0.066 6 0.297 0.341 0.396 1.034
1 8 1 6 1 0.074 0.057 0.066 0.197
 Example: Consistency Checking - 2
Step 2: Divide the elements of the vector of weighted
sums by the corresponding priority value.

1.803 0.593 3.040

1.034 0.341 3.032

0.197 0.066 2.985

Step 3: Compute the average of the values computed in

step 2 (max).

3.040 3.032 2.985

max 3.019
3
 Example: Consistency Checking - 3
Step 4: Compute the consistency index (CI).

max n 3.019 3
CI 0.010
n 1 3 1

Step 5: Compute the consistency ratio (CR).

CI 0.010
CR 0.017 0.10
RI 0.58
The degree of consistency exhibited in the pairwise comparison
matrix for comfort is acceptable.
Development of Priority Ranking
The overall priority for each decision
alternative is obtained by summing the
product of the criterion priority (i.e.,
weight) (with respect to the overall goal)
times the priority (i.e., preference) of the
decision alternative with respect to that
criterion.
Ranking these priority values, we will have
AHP ranking of the decision alternatives.
Example:
 Example: Priority Ranking 0A
Step 0A: Other pairwise comparison matrices

Comfort Car A Car B Car C Price Car A Car B Car C

Car A 1
Criterio 2 Price 8 Car Comfort
MPG A 1Style 1/3
Car B n
1/2 1 6 Car B 3 1
Car C Price
1/8 1/6 1 1 3 C
Car 2 4 2 2 1
MPG 1/3 1 1/4 1/4
MPG CarComfort
A Car B Car Style
1/2 C 4 1Car A 1/2Car B Car C
Car A 1
Style 1/4 1/21/6 Car
4 A 2 1 1 1/3 4
Car B 4 1 1/3 Car B 3 1 7
Car C 6 3 1 Car C 1/4 1/7 1
 Ranking of Criteria
Criterion Price MPG Comfort Style Criterion Price MPG Comfort Style
Price 1 3 2 2 Price 3/7 3/12 8/21 8/15
MPG 1/3 1 1/4 1/4 MPG 1/7 1/12 1/21 1/15
Comfort 1/2 4 1 1/2 Comfort 3/14 4/12 4/21 2/15
Style 1/2 4 2 1 Style 3/14 4/12 8/21 4/15
Column 7/3 12 21/4 15/4
total

Criterion

Criterion Price MPG Comfort Style Row Price 0.398

Average
0.398 MPG 0.085
Price 0.42 0.25 0.38 0.53 0,398
0.085 Comfort
MPG 0.143 0.083 0.048 0.067 0,085
Style
0.218
Comfort 0.214 0.33 0.19 0.133 0,218 0.218
0.299
Style 0.214 0.33 0.381 0.267 0,299
0 . 299 33
 Example: Priority Ranking 0B
Step 0B: Calculate priority vector for each matrix.
Price MPG Comfort Style Criterion

Car A 0.123 0.087 0.593 0.265

Price 0.398
Car B 0.320 0.274 0.341 0.655 MPG 0.085

Car C 0.557 0.639 0.066 0.080 Comfort 0.218
Style
0 . 299

Comfort Car A Car B Car C Row Avg.

0.593
Car A 0.615 0.632 0.533 0.593 0.341

Car B 0.308 0.316 0.400 0.341
0.066

Car C 0.077 0.053 0.067 0.066
Total 1.000
 Ranking of Criteria
Selecting a New Car
1.00

Comfort Price MPG Style

0.218 0.398 0.085 0.299

Car A 0.593 Car A 0.123 Car A 0.087 Car A 0.265

Car B 0.341
Car B 0.320

Car B 0.274

Car B 0.655


Car C 0.066 Car C 0.557 Car C 0.639 Car C 0.080

35
 Ranking of Alternative
Price MPG Comfort Style Criterion

Car A 0.123 0.087 0.593 0.265 Price 0.398

0.320 0.274 0.341 0.655 MPG 0.085
Car B
Car C 0.557 0.639 0.066 0.080 Comfort 0.218
Style
0 . 299
Price

Style
Comfort
MPG

Car A 0.123 0.087 0.593 0.265 0.398 0.265

0.085
Car B 0.320 0.274 0.341 0.655 x = 0.527
Car B is the
Car C 0.557 0.639 0.066 0.080 0.218 0.338
best alternative
0 . 299

Priority matrix
Criteria Weights 36
Complex decisions
Many levels of criteria and sub-criteria exists for
complex problems.

37
AHP Software:

Professional commercial software Expert Choice

developed by Expert Choice Inc. is available which
simplifies the implementation of the AHPs steps and
automates many of its computations

computations
sensitivity analysis
graphs, tables

38
More about AHP: Pros and Cons
AHP is technique for formalizing decision making such that
It is applicable when it is difficult to formulate criteria evaluations, i.e., it allows
qualitative evaluation as well as quantitative evaluation.
It is applicable for group decision making environments

However
There are hidden assumptions like consistency
Difficult to use when there are large number of evaluations Use GDSS
Use constraints to
eliminate some
alternatives

Difficult to add a new criterion or alternative Use cost/benefit

ratio if applicable

Difficult to take out an existing criterion or alternative, since

the best alternative might differ if the worst one is excluded.

39
Group Decision Making
The AHP allows group decision making, where group members can use their
experience, values and knowledge to break down a problem into a hierarchy
and solve. Doing so provides:

Understand the conflicting ideas in the organization and try to reach a

consensus.
Minimize dominance by a strong member of the group.
Members of the group may vote for the criteria to form the AHP tree.
(Overall priorities are determined by the weighted averages of the
priorities obtained from members of the group.)

However;
The GDSS does not replace all the requirements for group decision making.
Open meetings with the involvement of all members are still an asset.

40
 References
Al Harbi K.M.A.S. (1999), Application of AHP in Project Management, International
Journal of Project Management, 19, 19-27.

Haas R., Meixner, O., (2009) An Illustrated Guide to the Analytic Hierarchy Process,
Lecture Notes, Institute of Marketing & Innovation, University of Natural Resources
and http://www.boku.ac.at/mi/

Saaty, T.L., Vargas, L.G., (2001), Models, Methods, Concepts & Applications of the
Analytic Hierarchy Process, Kluwers Academic Publishers, Boston, USA.

41
Exercise

42
 Case
A motorist is using the AHP to choose a new car from three possible
models: an Arrow, a bestmobile and a commuter. The choice will be
based on just two attributes: cost and style. The motorist considers
that cost is weakly more important than style.

When asked to compare the costs of the cars, the motorist makes
the followig statements: on cost, the bestmobile is weakly preferred
to the arrow, but the arrow is weakly preferred to the commuter. Also
the bestmobile is extremely preferred to the commuter.

On style, the arrow is very strongly preferred to the bestmobile, but

the commuter is weakly preferred to the arrow. Also the commuter is
extremely preferred to the bestmobile.

a. Construct a hierarchy to represent the decision problem

b. Calculate the weight for each table in hierarchy and hence
determine which car should be purchased
c. Calculate the inconsistency ratios for the motorist comparisons of
the cars on (i) cost and (ii) style and interpret your results 43