Professional Documents
Culture Documents
Multiple Criteria Decision Analysis (MCDA)
Multiple Criteria Decision Analysis (MCDA)
Multiple Criteria Decision Analysis (MCDA)
(MCDA)
Multiple goals
Multiple alternatives to meet goals
Multiple criteria for alternatives
Multiple decision-makers with . . .
Diverse preferences
measures
Alternative 1 Alternative 2
Level 1: Criteria
4
Example: Car Selection
Objective
– Selecting a car
Criteria
– Style, Reliability, Fuel-economy
Alternatives
– Civic Coupe, Saturn Coupe, Ford Escort, Mazda
Miata
Hierarchy tree
S e le c t in g
a N ew C ar
max n
CI ;
n 1
RI ;
CI
CR .
RI
An inconsistency of 10% or less implies that the adjustment is small as compared to
the actual values of the eigenvector entries.
A CR as high as, say, 90% would mean that the pairwise judgments are just about
random and are completely untrustworthy! In this case, comparisons should be
repeated.
10
1
max n
CI ;
n 1
Weights determined RI ;
CI
1 0.5 3 CR .
RI
2 1 4
0.333 0.25 1.0
λmax=3.02
1
Consistency Ratio
The final step is to calculate the Consistency Ratio, CR by using the
table below, derived from Saaty’s book. The upper row is the order of
the random matrix, and the lower row is the corresponding index of
consistency for random judgments.
1
Ranking alternatives
Style Civic Saturn Escort Miata
Civic 1/1 1/4 2/1 1/6
Saturn 4/1 1/1 4/1 1/4
Escort 1/2 1/4 1/1 1/5
Miata 6/1 4/1 5/1 1/1
Eigenvector
.1160
.2470
.0600
.5770
1
Ranking alternatives
Reliability Civic Saturn Escort Miata
Civic 1/1 2/1 5/1 1/1 .3790
.2900
Saturn 1/2 1/1 3/1 2/1
.0740
Escort 1/5 1/3 1/1 1/4
.2570
Miata 1/1 1/2 4/1 1/1
1
Miles/gallon Normalized
113 1.0
1
S e le c t in g
a N ew C ar
1 .0
1
Ranking of alternatives
Style Reliability Fuel
Economy
Civic .1160 .3790 .3010 .3196
*
Saturn .2470 .2900 .2390 .5584
Escort .0600 .0740 .2120
.1220
Miata .5770 .2570 .2480
1
Handling Costs
1
Cost
Normalized
Cost Cost
MIATA $18K .333
CIVIC $12K .222
SATURN $15K .2778
ESCORT $9K .1667
2
Yaahp Software
1. Draw the graph:
2
2. Input the pairwise matrices:
In this example:
CR=CI/0.58=0.01/0.58=0.02 ,
0.02<0.1, so the evaluations are
consistent!
2
2
3. Adjust consistency:
2
2
4. Output the final results:
2
The Evaluation of System Risk of the Commercial Aircraft based on
Yaahp
2
MCDA
1.AHP
2.TOPSIS
3.DEA
2
Comparison Matrix
Given: Three apples of different
sizes.
Apple B S2 / S 1 S 2 / S2 S 2 / S3
Apple C S3 / S 1 S 3 / S2 S 3 / S3
2
Pairwise Comparisons
Size
Apple A 1 2 6
Apple B 1/2 1 3
3
0.8847+0.4423+0.1474=1.47
44
W=[0.8847/1.4744, 0.4423/1.4744, 0.1474/1.4744]’
=[0.6, 0.3, 0.1]’
3
Pairwise Comparisons
Size
Apple A 1 2 6 6/10 A
For X = (xij )
mn
xij
yij m 1 i m,1 j n
x
i 1
ij
2
Matrix Y yij m n
m
y
i 1
ij
2
1
0 yij 1
3
二 . Linear Scale Approach
If xj min xij 0
1 i m
yij x j
1 i m,1 j n
xij
Y yij m n
0 yij 1
3
三 . Range Approach
xj o xij 1 i m,1 j n
yij o
xj xj
xj 0 xij
yij 0 1 i m,1 j n
xj xj
Y yij m n 0 yij 1
3
Convert qualitative data to quantitative data
Very
Qualitative Index Very Bad Bad Good Excellent
Good
Positive Index 1 3 5 7 9
Reverse Index 9 7 5 3 1
3
Approach 1 : Evaluate the following four types of
airplanes using the linear weighed method.
Given the weight W (0.2, 0.1, 0.1, 0.1, 0.2, 0.3)T
Quantitative 1 3 5 7 9
Cost
C Speed Scope Load Reliab Sensit
( Million
type (Mach) (Km) ( Kg ) ility ivity
)
a1
2.0 1500 20000 5.5 5 9
a2 2.5 2700 18000 6.5 3 5
3
Linear Scale Approach
0.80 0.56 0.95 0.82 0.71 1.00
1.00 1.00 0.86 0.69 0.43 0.56
Y ( yij ) 46
0.72 0.74 1.00 1.00 1.00 0.78
0.88 0.67 0.95 0.90 0.71 0.56
u ( a* ) max ui u3 u (a3 )
1 i 4
a* a3
a3 a1 a4 a2
4
四 . TOPSIS(Technique for Order Preference by
Similarity to Ideal Solution)
v11 v12 v1n
v v v
V (vij ) mn 21 22 2n
vm1 vm 2 vmn
Determine the ideal solution and the negative-ideal
solution.
S {V1 ,V2 ,Vn } {(max j vij | i I ' ), (min j vij ) | i I "}
S {V1 , V2 ,Vn } {(min j vij | i I ' ), (max j vij ) | i I "}
I’ is associated with benefit criteria, and I’’ is associated
with cost criteria.
4
The relative closeness to the ideal solution:
n
d ( si , S ) d i (v
j 1
ij V 2
j ) , (i 1, 2, , m)
n
d ( si , S ) di ij j ) , (i 1, 2,, m)
( v
j 1
V 2
di
Ci
, (i 1, 2, , m)
di di
4
Example: Evaluate the following four types of
airplanes using TOPSIS approach.
Given the weight W (0.2, 0.1, 0.1, 0.1, 0.2, 0.3)T
Cost
C Speed Scope Load Reliab Sensit
( Million
type (Mach) (Km) ( Kg ) ility ivity
)
a1
2.0 1500 20000 5.5 5 9
a2 2.5 2700 18000 6.5 3 5
The =0.643
relative
,
closeness to the
=0.268 ,
ideal solution:
=0.613 , =0.312 ,
Using
d1* TOPSIS method
d 2* to make
d3* decision d 4*
d1 d 2 d 3 d 4
a1 a3 a4 a2
* *
C1Linear C
Scale
2 Approach C4*
a3 a1 a4 a2
a1 a3 a4 a2
4
Economy Versus the Environment
Table 1 1985-2002 normalized data
4
Introduction to Rough Set
47 4
a: salary; b: work condition; k: profit; l= Investment ; m: satisfaction
4
Conflict Analysis based on Rough Set in E-commerce
4
Conflict Analysis based on Rough Set in E-commerce
5
5