Sunteți pe pagina 1din 30

The Analytical Hierarchy Process (AHP)

Developed

by Thomas Saathy
A method to select the best one
To answer the question which one?
To buy a house from several houses
To select a university from a group
of universities
To choose the most qualified
candidate, etc.

2003 Thomson/South-Western

Slide 1

The Analytical Hierarchy Process (AHP)


AHP

is a method to generate a
numerical score to rank each decision
alternatives
The numerical score can be calculated
manually or can use excel

2003 Thomson/South-Western

Slide 2

The decision maker decides the


score for each alternative on a
criterion using pairwise comparison.

2003 Thomson/South-Western

Slide 3

In pairwise comparison, the decision


maker compares two alternatives
(a pair) according to one criterion
and indicates a preference.

2003 Thomson/South-Western

Slide 4

Preference Scale for Piecewise Comparison


Preference Level
Equally Preferred
Equally to moderately Preferred

Numerical value
1
2

Moderately Preferred

Moderately to Strongly
Preferred
Strongly Preferred
Strongly to very Strongly
Preferred
Very Strongly Preferred
Very Strongly to Extremely
Preferred
Extremely Preferred

2003 Thomson /South-Western

5
6
7
8
9

Slide 5

Let illustrate AHP by using an example.


PETRONAS SITE SELECTION PROBLEM

Suppose that PETRONAS has identified three


potential sites for its latest project near Alor
Setar (A), Bukit Merah (B) and Cyberjaya (C).
The company has identified four primary
criteria on which it will compare the sites
namely:
2003 Thomson/South-Western

Slide 6

The four criteria:

The customer market (overall market size, and


population at different age level)
Income level
Infrastructure (including utilities, and roads)
Transportation (proximity to highway for supplier
deliveries and customer access)

The overall objective of the company is to


select the best site

2003 Thomson/South-Western

Slide 7

1. Comparison matrix for customer market:


CUSTOMER MARKET
A

1/3

1/5

1/2

2003 Thomson/South-Western

Slide 8

2. Comparison matrix for income level:


INCOME LEVEL
A

1/3

1/6

1/9

2003 Thomson/South-Western

Slide 9

3. Comparison matrix for infrastructure:


INFRASTRUCTURE
A

1/3

1/7

2003 Thomson/South-Western

Slide 10

4. Comparison matrix for transportation:


TRANSPORTATION
A

1/3

1/2

1/4

2003 Thomson/South-Western

Slide 11

Let us do synthesization: decision alternatives


are prioritized within each criterion.
STEP 1

CUSTOMER MARKET
A

1/3

1/5

1/2

Total

11/6

16/5

2003 Thomson/South-Western

Slide 12

Next, we divide each value in a column by its


corresponding column sum. This results in a
normalized matrix as follows:
STEP 2
CUSTOMER MARKET
A

6/11

3/9

5/8

2/11

1/9

1/16

3/11

5/9

5/16

2003 Thomson/South-Western

Slide 13

Next, change them to decimal and find the


average for each row.
STEP 3

CUSTOMER MARKET

Row
average

0.5455

0.3333

0.6250

0.5012

0.1818

0.1111

0.0625

0.1185

0.2727

0.5556

0.3125

0.3803

2003 Thomson/South-Western

Slide 14

Thus, based on customer market, the sites are


ranked as follows:

0.5012 (A)
0.1185 (B)

0.5012 (A)
OR

0.3803 (C)
2003 Thomson/South-Western

0.3803 (C)
0.1185 (B)
Slide 15

We have to do the same process to the other 3


criteria. The results are as follows:

Income
Level

Infrastructure Transportation

0.2819
(A)

0.1780 (A)

0.1561 (A)

0.0598
(B)

0.6850 (B)

0.6196 (B)

0.6583
(C)

0.1360 (C)

0.2243 (C)

2003 Thomson/South-Western

Slide 16

LET COMBINE ALL 4 CRITERIA TOGETHER.

CRITERIA (CALL THIS MATRIX M)

MARKET

INCOME
LEVEL

0.5012

0.2819

0.1790

0.1561

0.1185

0.0598

0.6850

0.6196

0.3803

0.6583

0.1360

0.2243

2003 Thomson/South-Western

INFRATRANS
STRUCTURE PORTATION

Slide 17

NEXT, WE WANT TO RANK THE CRITERIA.

2003 Thomson/South-Western

Slide 18

WELL ASK THE DM TO FILL THIS TABLE.

CM

IL

IS

TP

CM

1/5

IL

IS

1/3

1/9

TP

1/4

1/7

1/2

2003 Thomson/South-Western

Slide 19

WELL DO THE SAME PROCESS AS BEFORE:


NOW, ADD THE VALUES IN EACH COLUMN

CM

IL

IS

TP

CM

1/5

IL

IS

1/3

1/9

TP

1/4

1/7

1/2

TOT

79/12

458/315

27/2

14

2003 Thomson/South-Western

Slide 20

NEXT, DIVIDE EACH ENTRY WITH RESPECTIVE


COLUMN SUM.

CM

IL

IS

TP

CM

12/79

315/2290

6/27

4/14

IL

18/27

7/14

IS

2/27

2/14

TP

1/27

1/14

2003 Thomson/South-Western

Slide 21

NEXT, WRITE EACH ENTRY INTO DECIMAL FORM.

CM

IL

IS

TP

CM

0.1519

0.1375

0.2222

0.2857

IL

0.6667

0.5000

IS

0.0741

0.1429

TP

0.0370

0.0714

2003 Thomson/South-Western

Slide 22

NEXT, TAKE AVERAGE FOR EACH ROW.

CM

IL

IS

CM 0.1519

0.1375

0.2222

Average
Row
0.2857 0.1993

IL

0.6667

0.5000

IS

0.0741

0.1429

TP

0.0370

0.0714

2003 Thomson/South-Western

TP

Slide 23

THOSE NUMBERS REFLECT THE RANK FOR THE


CRITERIA. WE ARE SUPPOSE TO HAVE:

ROW AVERAGES(CALL
THIS MATRIX N)

0.1993 (CM)
0.6535 (IL)
0.0860 (IS)
0.0612(TP)
2003 Thomson/South-Western

Slide 24

NOW WE ARE READY TO DEVELOP AN OVERALL


RANKING. WHAT WE HAVE TO DO IS TO MULTIPLY
MATRIX M AND N.

0.1993

0.5012 0.2819 0.1790 0.1561

0.6535
0.1185 0.0598 0.6850 0.6196

0
.
0860
0.3803 0.6583 0.1360 0.2243

0.0612

2003 Thomson/South-Western

Slide 25

WE ARE SUPPOSE TO GET:

Score for site A


Score for site B
Score for site C

0.3091

0.1595
0.5314

2003 Thomson/South-Western

Slide 26

REARRANGE ACCORDING TO SCORE:


SCORE FOR SITE C IS 0.5314
SCORE FOR SITE A IS 0.3091
SCORE FOR SITE B IS 0.1595

THUS, SITE C (CYBERJAYA) IS THE BEST CHOICE.

2003 Thomson/South-Western

Slide 27

Make your own summary of steps

2003 Thomson/South-Western

Slide 28

Assignment: Use excel to solve


Petronas Site Selection Problem.

2003 Thomson/South-Western

Slide 29

Food of thought
The man ignorant of mathematics will
be increasingly limited in his grasp of
the main forces of civilization.

2003 Thomson/South-Western

Slide 30

S-ar putea să vă placă și