You are on page 1of 40

Analytic Hierarchy Process

AHP is a decision-making technique.


It provides a comprehensive framework for making
multicriteria decisions by organizing problems into a
hierarchical structure.

It is a systematic procedure for representing the


elements of any problem, hierarchically.

It organizes the basic rationality by decomposing a


general decision problem in a hierarchical fashion
into subproblems that can be easily comprehended
and evaluated;
AHP

determining the priorities of the elements at


each level of the decision hierarchy through
a series of pair-wise comparison judgments
to express the relative strength or intensity
of impact of each element in the hierarchy;
and synthesizing the priorities to determine
the overall priorities of the decision
alternatives.
AHP
Developed by Thomas L Saaty.
Designed to solve complex multi criteria
decision problems.
AHP requires the decision maker to provide
judgment about relative importance of each
criterion and specify a preference for each
decision alternatives using each criteria
The output of AHP is a prioritized ranking of
each decision alternative based on the overall
preferences expressed by decision maker.
The Standard form of the AHP Hierarchy
.

Level 1 : General objective of


Focus the decision problem

Level II : Decision Decision Decision


attributes Attribute 1 Attribute 2 Attribute n

Level II : More detailed More detailed More detailed More detailed More detailed
subattributes subattribute subattribute subattribute subattribute subattribute

More detailed
subattribute
Level K : Decision Decision Decision
alternatives Attribute 1 Attribute 1 Attribute 1
The Main Operational Steps of AHP
A brief description of the AHP operational steps is as follows :
1. Define the problem.
2. Structure a hierarchy representing the problem. Arrange
goals, attribute, criteria, subcriteria, issues, activities,
alternatives, and so on, in a hierarchy.
3. Perform pair-wise comparison judgments on elements at each
level of the hierarchy with respect to another element higher
up the hierarchy. This process produces a series of pair-wise
comparison matrices at each level of the hierarchy.
4. Compute the local weighting of the elements of each level with
respect to an element higher up the hierarchy.
5. Use hierarchical composition to combine the weighting to
obtain the global weightings for the alternatives.
6. Check the model and repeat any part as required.
The Structure of AHP Hierarchy

The top level, called the focus, consists of only one


element- the board, overall objective. Subsequent
levels may each have several elements, although
their number is very small-between 5 and 9.
Because the elements in one level are to be
compared with one another against a criterion in the
next higher level, the elements in each level must be
of the same order of magnitude.
Example: Car Purchase
Car options : Honda Accord, Saturn and
Chevrolet
Criteria relevant:
I. Price
II.Miles per gallon
III.Comfort
IV.Style
Developing Hierarchy

Overall Goal
Select the best car

Price MPG Comfort Style

Accord
Accord Accord Accord Accord
Saturn
Saturn Saturn Saturn Saturn
Chevrolet
Chevrolet Chevrolet Chevrolet Chevrolet
Rating Scales for Verbal Judgment
.
Judgments Rating Reciprocal Judgment Rating

Extremely preferred 9 Extremely less preferred 1/9

Very strongly preferred 7 Very strongly less preferred 1/7

Strongly preferred 5 Strongly less preferred 1/5

Moderately preferred 3 Moderately less preferred 1/3

Equally preferred 1 Equally less preferred 1

And in between preferences And in between preferences


(8, 6, 4 and 2) (0.13, 0.17, 0.25, and 0.5)
Information
Characteristics Accord saturn Chevrolet
Price 13 100 11200 9500
Color Black Red Blue
Miles per gallon 19 23 28
Interior Deluxe Above average standard
Body type 4 door mid size 2 door sport 2 door
compact
Sound system AF?FM, CD AM/FM AM/FM
Rating Scales for Verbal Judgment
.
Judgments Rating Reciprocal Judgment Rating

Extremely preferred 9 Extremely less preferred 1/9

Very strongly preferred 7 Very strongly less preferred 1/7

Strongly preferred 5 Strongly less preferred 1/5

Moderately preferred 3 Moderately less preferred 1/3

Equally preferred 1 Equally less preferred 1

And in between preferences And in between preferences


(8, 6, 4 and 2) (0.13, 0.17, 0.25, and 0.5)
Pair wise comparison matrix
Establish the priorities for four criteria

Price MPG Comfort Style


Price 1 3 2 2
MPG 1/3 1 1/4 1/4
Comfort 1/2 4 1 1/2
Style 1/2 4 2 1
Synthesization
Sum the values in each column of the pair
wise comparison matrix
Normalized matrix: divide each element in
the pairwise comparison matrix by its column
total
Compute the average of the elements in each
row of the normalized pair wise comparison
matrix. These averages provides the priorities
for the criteria
Price
Price MPG
MPG Comfort
Comfort Style
Style
Price
Price 11 33 22 22
MPG
MPG 1/3
1/3 11 1/4
1/4 1/4
1/4
Comfort
Comfort 1/2
1/2 44 11 1/2
1/2
Style
Style 1/2
1/2 44 22 11
Total 2.333 12.00 5.25 3.75
Price
Price MPG
MPG Comfort
Comfort Style
Style
Price 1 3 2 2
Price 0.429 0.25 0.381 0.533
MPG 1/3 1 1/4 1/4
MPG 0.143 0.083 0.048 0.067
Comfort 1/2 4 1 1/2
Comfort 0.214 0.333 0.190 0.133
Style 1/2 4 2 1
Style 0.214 0.333 0.381 0.267
Price
Price MPG
MPG Comfort
Comfort Style Style Priority
Price 1 3 2 2
Price 0.429 0.25 0.381 0.533 0.398
MPG 1/3 1 1/4 1/4
MPG 0.143 0.083 0.048 0.067 0.085
Comfort 1/2 4 1 1/2
Comfort 0.214 0.333 0.190 0.133 0.218
Style 1/2 4 2 1
Style 0.214 0.333 0.381 0.267 0.299
Consistency Test
Examine the consistency of the pairwise
judgments provided.
Consistency ratio provides a measure of
consistency.
Consistency ratio greater than 0.10 indicates
an inconsistency of the pair wise comparison.
Consistency ratio less than 0.1 is considered
reasonable.
Test for Consistency
1.Multiply each value in the first column of the
pairwise comparison by the priority of the first item
and so on.
2. Divide the element of the weighted sum
vector in the step 1 by the corresponding
priority of each criterion
Price = 1.687/ 0.398 = 4.235
MPG = 0.347/ 0.085 = 4.077
Comfort = 0.907/0.218 = 4.163
Style = 1.274/ 0.299 = 4.264
Test for Consistency
Price
Price MPG
MPG ComfortStyleStyle Priority
Comfort
Price 1 3 2 2
MPG 1/3 1 1/4 1/4
Price 1 3 2 2 0.398 1.687
Comfort 1/2 4 1 1/2
Style
MPG 1/2
1/3 14 1/4 2 1/4 1 0.085 0.347

Comfort 1/2 4 1 1/2 0.218 0.907

Style 1/2 4 2 1 0.299 1.274

Priority 0.398 0.085 0.218 0.299


Test for Consistency

2. Divide the element of the weighted sum


vector in the step 1 by the corresponding
priority of each criterion
Price = 1.687/ 0.398 = 4.235
MPG = 0.347/ 0.085 = 4.077
Comfort = 0.907/0.218 = 4.163
Style = 1.274/ 0.299 = 4.264
Consistency Index
3. Compute average = 4.185 ( max )

4. Compute CI = ( max - n)/ ( n-1)


n= number of items being compared

CI = ( 4.185 4 )/3 = 0.0616

5. Compute the consistency ratio


CR = CI/ RI = 0.0616/ 0.90 = 0.068
CR ration of 0.10 or less is considered acceptable
RI is the consistency index of randomly
generated pairwise comparisons matrix
n 3 4 5 6 7 8

RI 0.58 0.90 1.12 1.24 1.32 1.41


Pair wise comparison showing
preferences for the cars
PRICE Accord Saturn Chevrolet
Accord 1 1/3 1/4
Saturn 3 1 1/2
Chevrolet 4 2 1

MPG Accord Saturn Chevrolet


Accord 1 1/4 1/6
Saturn 4 1 1/3
Chevrolet 6 3 1
Pair wise comparison showing
preferences for the cars
COMFORT Accord Saturn Chevrolet
Accord 1 2 8
Saturn 1/2 1 6
Chevrolet 1/8 1/6 1

STYLE Accord Saturn Chevrolet


Accord 1 1/3 4
Saturn 3 1 7
Chevrolet 1/4 1/7 1
Priorities for each car using each
Criterion
PRICE MPG (0.085) COMFORT STYLE
(0.398) ( 0.218) ( 0.299

ACCORD 0.123 0.087 0.593 0.0265

SATURN 0.320 0.274 0.341 0.656

Chevrolet 0.557 0.639 0.065 0.080


Overall Priority
Accord :
= 0.398*(0.123) + 0.085*( 0.593) +
0.299*(0.265) = 0.265

Saturn = 0.421

Chevrolet = 0.314
Rating Scales for Verbal Judgment
.
Judgments Rating Reciprocal Judgment Rating

Extremely preferred 9 Extremely less preferred 1/9

Very strongly preferred 7 Very strongly less preferred 1/7

Strongly preferred 5 Strongly less preferred 1/5

Moderately preferred 3 Moderately less preferred 1/3

Equally preferred 1 Equally less preferred 1

And in between preferences And in between preferences


(8, 6, 4 and 2) (0.13, 0.17, 0.25, and 0.5)
Example

Evaluating / Weighing Guns in a Weapon


System
Objective
Evaluating / Weighing a Gun in a Weapon
System
Criteria / Attributes
Range, Mobility, Survivability
Alternatives
Gun 1, Gun 2, Gun 3, Gun 4
Hierarchical Tree
.

Evaluation of WEI

Level I Attack Defence

Level II
Artillery Armour Infantry

Level III Range Mobility Survivability


Equipment 1
Equipment 2
Equipment 3
Assigning Weightage to
Characteristics
Pair-wise relative weightage
1: Equal, 3:Moderate, 5: Strong, 7: Very strong,
9: Extreme
Mobility Survivability Range
Mobility 1/1 1/2 3/1
Survivability 2/1 1/1 4/1
Range 1/3 1/4 1/1
Ranking of Weightages

Eigenvector A x x

Iterate
1. Take successive squared powers of matrix
2. Normalize the row sums

Until difference between successive row sums is less


than a pre-specified value.
.

1 0.5 3 3.0 1.75 8.0


2 1 4 squared 5.3332 3.0 14.0

0.333 0.25 1.0 1.1666 0.6667 3.0

Row sums Normalized


Row sums
12.76 0.3194
22.3332 0.5595
4.8333 0.1211
39.9165 1.0
0.3196
New iteration gives normalized row sum
0.5584

0.1220

0.3194 0.3196 0.0002


Difference is :
0.5595 0.5584 0.0011

0.1211 0.1220 0.0009
Weightage

Mobility 0.3196
Survivability 0.5584
Range 0.1220
Ranking Guns
..

Mobility Gun 1 Gun 2 Gun 3 Gun 4 Eigenvector

Gun 1 1/1 1/4 4/1 1/6


0.1160
0.2470
Gun 2 4/1 1/1 4/1 1/4

Gun 3 1/4 1/4 1/1 1/5 0.0600
Gun 4 6/1 4/1 5/1 1/1
0.5770
Ranking of Guns
..

Surviva- Gun 1 Gun 2 Gun 3 Gun 4


bility 0.3790
Gun 1 1/1 2/1 5/1 1/1 0.2900
Gun 2 1/2 1/1 3/1 2/1
Gun 3 1/5 1/3 1/1 1/4 0.0740

Gun 4 1/1 1/2 4/1 1/1 0.2570
.

Effective Range Normalized


(meters)
Effectiveness Gun 1 340 0.3010
(quantitative
information) Gun 2 270 0.2390

Gun 3 240 0.2120

Gun 4 280 0.2480

1130 1.0
.

Weightage
1.0

Mobility Reliability Range


0.3196 0.5584 0.1220

-Gun1 0.1160 -Gun1 0.3790 -Gun1 0.3010


-Gun2 0.2470 -Gun2 0.2900 -Gun2 0.2390
-Gun3 0.0600 -Gun3 0.0740 -Gun3 0.2120
-Gun4 0.5770 -Gun4 0.2570 -Gun4 0.2480
Weights Effectiveness of each Gun

Mobility Survivability Range

Gun1 0.1160 0.3790 0.3010 0.3060


0.3196
Gun 2 0.2470 0.2900 0.2390 0.2720
* 0.5584
Gun3 0.600 0.0740 0.2120 0.940
0.1220
Gun 4 0.5770 0.2570 0.2480 0 .3280
Cost vs. Benefits
.

Cost Normalized Cost/Benefits


Cost Ratio
Gun 1 Rs. 18 K 0.333 0.9840

Gun 2 Rs. 12 K 0.222 1.3771

Gun 3 Rs. 15 K 0.2778 0.9791

Gun 4 Rs. 9 K 0.1667 0.5639

You might also like