You are on page 1of 17

Analytical Hierarchy Process

Rini Novrianti Sutardjo Tui


Decision Making

Deductive
approach System
considers problem as a approach considers
network and breaks it
problem as a whole entity.
down into links and
cycles.
Attitude in Decision Making

Dynamic
Field

Instinct Reason –
– Drive Impulse
Analytical Hierarchy Process

Analytical Hierarchy Process (AHP) is a method which


breaks complicated problem into small components and
arranges them into hierarchic sequences, provides
quantitative judgment, and determines component
which has the highest priority.
Principles of Analytical
Thinking

Developing
hierarchy

Setting
Logical
priorities
consistency
Hierarchies

Hierarchy
Identifying, grouping,
arranging

Structural Functional
complex system decomposed complex
structured into their system into their
constituent parts constituent parts
according to their according to their
structural properties. essential relationships.
Why Analytical Hierarchy
Process?

Unity: Single model, flexible, understandable


Complexity: Unites two approaches

Interdependence: Not pushing linear thinking


Hierarchy Structuring: Grouping

Measurement: Measures intangibles and sets priority


Why Analytical Hierarchy
Process?

Consistency: Tracks down consistency


Synthesis: Towards estimated desirability

Trade-offs: Considers priority

Judgment & Consensus: Not pushing consensus

Process Repetition: Improves by repetition


Some Application of the Hierarchy

Financial
Decisions Personal &
Business Domestic
Decisions Decisions

Public Policy
Estimating & Decisions
Predicting Economic
Policies
Scale for Pairwise Comparisons

Source: Saaty (1980)


Example 1

Good Choice for a Sport Car

Satisfaction of Freedom
Salary Prestige Basic Needs Comfort Other Needs from Worry

Porsche Mercedes Aston Martin Audi Lamborghini


Calculation
To choose a sport car on the basis of comfort

Comfort Porsche Mercedes Aston Martin Audi Lamborghini

Porsche 1 1/4 1/2 1/3 1/2

Mercedes 4 1 3 2 4

Aston Martin 2 1/3 1 3 4

Audi 3 1/2 1/3 1 2

Lamborghini 2 1/4 1/4 1/2 1

Total 12 7/3 61/12 41/6 23/2


Normal Matrix

Aston Row Av. Row


Comfort Porsche Mercedes
Martin Audi Lamborghini
Sum Sum

Porsche 1/12 3/28 6/61 2/41 1/23 0.38 0.076

Mercedes 1/3 3/7 36/61 12/41 8/23 1.99 0.398

Aston Martin 1/6 1/7 12/61 18/41 8/23 1.29 0.258

Audi 1/4 3/14 4/61 6/41 4/23 0.85 0.170

Lamborghini 1/6 3/28 3/61 3/41 2/23 0.48 0.096

Total 1
Consistency Index (CI)

max  n
Consistency Index=
n 1

row _ total
max 
average _ row _ sum
Consistency Index

Porsche Mercedes
Aston Audi
Comfort (0.076) (0.398)
Martin Lamborghini Row Total
(0.258) (0.170) (0.096)

Porsche 0.076 0.0995 0.129 0.057 0.048 0.4095

Mercedes 0.304 0.398 0.774 0.340 0.384 2.2000

Aston
0.152 0.133 0.258 0.510 0.384 1.4370
Martin
Audi 0.228 0.199 0.086 0.170 0.192 0.8750

Lamborghini 0.152 0.0995 0.0645 0.085 0.096 0.4970

0.4095 0.076 5.388 5.388  5.528  5.569  5.147  5.177 26.809


max    5.3618
2.2000 0.398 5.528 5 5
1.4370   0.258  5.569
0.8750 0.170 5.147
0.4970 0.096 5.177 max  n
5.3618  5
      CI    0.09
n 1 4
Consistency Ratio

CI
Consistency Ratio=
RCstd
Maximum CR value allowed is 10%

Size of Matrix 1 2 3 4 5 6 7 8 9 10
Random Consistency 0.00 0.00 0.52 0.89 1.11 1.25 1.35 1.40 1.45 1.49

To choose a sport car on the basis of comfort

CI 0.09
CR    0.08
RCstd _ n5 1.11

You might also like