You are on page 1of 4

30-Jan-18

Objective

The objective of this lecture is as follows

► The purpose of this Lecture is to facilitate you to understand the concept


of decision making when ideas feelings & emotions affecting the decision..

Module-I: Decision Theory The topics will be covered in this lecture are mentioned below.

Lecture: 5 & 6 – Analytic Hierarchy Process ► Analytic hierarchy process

► Numerical Problem on AHP

► Consistency of Comparison matrix

Decision making with emotions Analytic Hierarchy Process (AHP)

Analytic Hierarchy Process (AHP) Applications

► AHP is designed for the situations in which ideas feelings, emotions, ► Wide range of applications are there in personal life as well as
affecting the decision process are qualified to provide a numeric scale for professional life
prioritizing the alternatives.
► Selection a Car/ House for purchaseing
► AHP is a structured technique for organizing & analyzing complex
decisions. ► Decision upon a place to visit for vacation

► It was developed by Thomas L Saaty in the 1970’s ► Decision upon a MBA programme after graduation

► AHP algorithm is basically composed of 2 steps ► Finalization of Contractor after competitive Bidding etc.
► Determine the relative weight of the decision criteria

► Determine the relative ranking of alternatives.

► Both qualitative & Quantitative information can be compared.

Analytic Hierarchy Process (AHP) Analytic Hierarchy Process (AHP)

Problem-01 Solution-01
Purchasing a New
Car
► You are planning to purchase one new car. You are looking for three
quality of the car i.e. Style, Cost & Fuel efficiency. After preliminary survey
you have chosen 4 options. Those are- Tata Tigor, Maruti Suzuki Baleno,
Hyundai i20 & Ford Fiesta. Which car you will buy finally?
Style Cost Fuel Economy

Maruti Suzuli
Tata Tiogr Hyundai i20 Ford Fiesta
Baleno

1
30-Jan-18

AHP AHP

Ranking Scale for Criteria & Alternatives Ranking Criteria

Scale Definition
Value
Style Cost Fuel
1 Equal importance Efficiency
3 Weak importance of one over another Style 1 1/2 3
5 Essential or strong importance Cost 2 1 4
7 Very strong importance
Fuel Efficiency 1/3 1/4 1
9 Absolute importance
2, 4, 6, 8 For the above two judgments of the intermediate state
corresponds to the value of the scaling

► Ref: T.L.Saaty.The Analytic hierarchy process. McGraw-Hill International


Book Company, New York, 1980

AHP AHP

Ranking of priorities Ranking of priorities

► Consider [AX = λmaxX] where


► A is the comparison matrix of size nxn, for n criteria, also called the

priority matrix
Normalized Row Average
► X is the Eigen vector of size nx1, also called the priority vector Column Sums

► λmax is the Eigen value. 1 0.5 3 0.3 0.28 0.37 0.32


► A= 2 1 4 = 0.6 0.57 0.51 X= 0.56
► To find the ranking o priorities. Namely the Eigen vector X 0.33 0.25 1 0.1 0.15 0.12 0.12
1. Normalize the column entries by dividing each entry by the sum of the
column Column Sum 3.33 1.75 8.00 1 1 1 Priority Vector
2. Take the overall row average .

AHP AHP

Ranking of priorities Checking the Consistency

► The next stage is to calculate a consistency ratio (CR) to measure how


consistent the judgements have been relative to large samples of purely
random judgements.
Purchasing a New
Car
1.00 ► AHP evaluations are based on assumptions that the decision maker is
rational i.e. if A is preferred to B & B is preferred to C the A is preferred to
C
Style Cost Fuel Economy
0.32 0.56 0.12 ► If the CR is > 0.1 then it is called in-consistent, The exercise to be
repeated.
► If the value is ≤ 0.1 then it is consistent.
► Style 0.32
► Cost 0.56
► Fuel Efficiency 0.12

2
30-Jan-18

AHP AHP

Calculation of Consistency Ratio Calculation of Consistency Ratio

► The next stage is to calculate λmax So as to lead the consistency index ► Consistency Index CI = (λmax-n)/ (n-1) = (3.04-3)/(3-1) = 0.02
and the consistency ratio
► Consider [AX = λmaxX] where X is Eigen vector ► Now CR = CI/ RI where RI is randomness index.

► CI =0.02
A X AX X ► n=3
1 0.5 3 0.32 0.98 0.32 ► RI = 0.58 (from table)
► 2 1 4 x 0.56 = 1.68 =λmax 0.56
0.33 0.25 1 0.12 0.36 0.12 ► So, CR = CI/RI = 0.02/0.58 = 0.034

► CR ≤ 0.1 indicates sufficient consistency for decision.


► λmax = Average (0.98/ 0.32, 1.68/0.56, 0.36/0.12) = 3.04

AHP AHP

Randomness Index (RI) value of 1-10 rank matrix Ranking Alternatives Level-3

Priority Vector
Style Tigor Baleno I20 Fiesta
Tigor 1 1/4 4 1/6 0.13
X= 0.24
Rank of matrix RI value Rank of matrix RI value Baleno 4 1 4 1/4
1 0.00 6 1.24 I20 1/4 1/4 1 1/5 0.07
2 0.00 7 1.32
3 0.58 8 1.41
Fiesta 6 4 5 1 0.56
4 0.90 9 1.45
5 1.12 10 1.49 Cost Tigor Baleno I20 Fiesta
Tigor 1 2 5 1 0.38
Baleno 1/2 1 3 2 X= 0.29
0.07
I20 1/5 1/3 1 1/4
► Ref: L.G.Vargas,An overview of the analytic hierarchy process and its 0.26
application. European Journal of Operational Research, 1990, 48:2. Fiesta 1 1/2 4 1

AHP AHP

Ranking Alternatives Level-3 Calculation of the total order to determine the final decision
Fuel Miles/ Priority
Economy Gallon Vector 0.13 0.38 0.30 0.32 0.29
0.24 0.29 0.24 x = 0.28
Tigor 34 0.3 ►
0.07 0.07 0.21 0.56 0.08
Baleno 27 0.24
0.56 0.26 0.25 0.12 0.35
I20 24 0.21
Fiesta 28 0.25
Total 113 1.00 ► As ford fiesta is getting the highest value in hierarchy
chart
Note: Since fuel economy is a quantitative measure, fuel consumptions
ratios can be used to determine the relative ranking of the alternatives.
► You should buy ‘Ford Fiesta’

3
30-Jan-18

AHP

Pros

► It allows multi criteria decision making


► It is applicable when it is difficult to formulate criteria evaluations i.e. it
allows qualitative as well as quantitative evaluation
► It is applicable for group decision making environments

Cons

► There are hidden assumptions like consistency. Repeating evaluation is


cumbersome.
Questions…
► Difficult to use when the number of criteria or alternative is high i.e.more
than 7
► Difficult to add a new criteria or alternative
► Difficult to take out an exiting criteria or alternative, since the best
alternative might differ if the worst one is excluded.

THANK YOU…

You might also like