IME634: Management Decision

Analysis
Raghu Nandan Sengupta
Industrial & Management Department
Indian Institute of Technology Kanpur

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 1

Technique for Order Preference by Similarity to
Ideal Solution (TOPSIS)
Technique for Order Preference by Similarity to
Ideal Solution (TOPSIS) method was developed
for the INTEGRated Human Exploration Mission
Simulation FaciliTY (INTEGRITY) project in the
Johnson Space Centre to assess the priority of a
set of human spaceflight mission simulators

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 2

 TOPSIS (contd…)
One assumes utility function is monotonic, in the
sense the more/less you get more/less you want
The basic premise based on which TOPSIS works
is the fact that selected alternatives should have
the shortest distance from the positive-ideal
solution, and the farthest distance from the
negative-ideal solution
TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 3
 TOPSIS (contd…)
 Choose positive ideal solution (PIS) of the original ranking
problem
 Choose negative ideal solution (NIS) of the original ranking
problem
 Find distances from each decisions/alternatives, 𝐴𝑖 , 𝑖 =
1, ⋯ , 𝑚 to PIS, which is given as 𝑑 𝐴𝑖 , 𝑃𝐼𝑆
 Find distances from each decisions/alternatives, 𝐴𝑖 , 𝑖 =
1, ⋯ , 𝑚 to NIS which is given as 𝑑 𝐴𝑖 , 𝑁𝐼𝑆

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 4

 TOPSIS (contd…)
Euclidean distance measure is used and we ensure our
main motivation is to minimize the dispersion
𝑑 𝐴𝑖 ,𝑁𝐼𝑆 𝐴𝑖 −𝑁𝐼𝑆 2
Calculate 𝑟𝑖 = =
𝑑 𝐴𝑖 ,𝑁𝐼𝑆 +𝑑 𝐴𝑖 ,𝑃𝐼𝑆 𝐴𝑖 −𝑁𝐼𝑆 2 + 𝐴𝑖 −𝑃𝐼𝑆 2
The basic premise being Euclidean distance portrays the
concept of utility function, 𝑈 𝑊 which is quadratic

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 5

 TOPSIS (contd…)

Minimizing 𝑈 𝑊 ensures
minimizing dispersion, i.e.,
minimization of 𝑉𝑎𝑟 𝑋
One ranks the ratios, 𝑟𝑖 , to get the
best alternative
TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 6
 Algorithm for TOPSIS
Assume decisions/alternatives as 𝐴𝑖 , 𝑖 =
1, ⋯ , 𝑚
Assume attributes/decision criteria/goals are
𝐶𝑗 , 𝑗 = 1, ⋯ , 𝑛
We state the pseudo-codes for the working
principle of TOPSIS

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 7

 Algorithm for TOPSIS (contd…)
1: DEFINE: 𝑿𝑚×𝑛 (matrix consisting of priority scores assigned to decisions/alternatives),𝐴𝑖 ,based on attributes/decision criteria/goals,𝐶𝑗 ;𝑤𝑗 (weight for the
𝑥𝑖,𝑗
attributes/decision criteria/goals) such that σ𝑛𝑗=1 𝑤𝑗 = 1;𝑩 (benefit matrix);𝑪 (cost matrix);𝑟𝑖,𝑗 = ; 𝐴𝐼𝑆 = 𝑣1+ , ⋯ , 𝑣𝑚
+ =
σ𝑚 2
𝑖=1 𝑥𝑖,𝑗

max 𝑣𝑖,𝑗 𝑗 ∈ 𝑩 , min 𝑣𝑖,𝑗 𝑗 ∈ 𝑪 (negative ideal solution);𝑃𝐼𝑆 = 𝑣1− , ⋯ , 𝑣𝑚

− = min 𝑣𝑖,𝑗 𝑗 ∈ 𝑩 , max 𝑣𝑖,𝑗 𝑗 ∈ 𝑪 (positive ideal solution);𝑆𝑖+ =
∀𝑖 ∀𝑖 ∀𝑖 ∀𝑖
2 2 𝑆𝑖−
σ𝑛𝑗=1 𝑣𝑖,𝑗 − 𝑣𝑗+ ;𝑆𝑖− = σ𝑛𝑗=1 𝑣𝑖,𝑗 − 𝑣𝑗− ;𝑇𝑖 = + (relative closeness);𝑀 = 𝑆𝑖+ , 𝑆𝑖− (separation measure). Here 𝑖 = 1, ⋯ , 𝑚 and 𝑗 = 1, ⋯ , 𝑛
𝑆𝑖 +𝑆𝑖−

2: INPUT: 𝑿𝑚×𝑛 (matrix consisting of priority scores assigned to decisions/alternatives),𝐴𝑖 ,based on attributes/decision criteria/goals,𝐶𝑗 ;𝑤𝑗 (weight for the
attributes/decision criteria/goals) such that σ𝑛𝑗=1 𝑤𝑗 = 1;𝑩 (benefit matrix);𝑪 (cost matrix). Here 𝑖 = 1, ⋯ , 𝑚 and 𝑗 = 1, ⋯ , 𝑛
3: START if: 𝑖 = 1: 𝑚
4: START if: 𝑗 = 1: 𝑛
𝑥𝑖,𝑗
5: CALCULATE: 𝑟𝑖,𝑗 = ;𝑣𝑖,𝑗 = 𝑤𝑗 𝑟𝑖,𝑗 where 𝑖 = 1, ⋯ , 𝑚 and 𝑗 = 1, ⋯ , 𝑛
σ𝑚 2
𝑖=1 𝑥𝑖,𝑗

6: END if
7: END if
8: CALCULATE: 𝑣𝑖+ ; 𝑣𝑖− ;𝐴𝐼𝑆;𝑃𝐼𝑆;𝑆𝑖+ ;𝑆𝑖− ; 𝑀;𝑇𝑖
9: REPORT: 𝐴𝐼𝑆;𝑃𝐼𝑆;𝑀;𝑇𝑖
10: END

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 8

 TOPSIS (contd…): Distance measure
 The Euclidean distance between vector/points 𝒙 = 𝑥1 , ⋯ , 𝑥𝑛
1
and 𝒚 = 𝑦1 , ⋯ , 𝑦𝑛 is σ𝑛𝑖=1 𝑥𝑖 − 𝑦𝑖 2 2
 The 𝐿1 norm or Manhattan distance between vector/points 𝒙 =
𝑥1 , ⋯ , 𝑥𝑛 and 𝒚 = 𝑦1 , ⋯ , 𝑦𝑛 is σ𝑛𝑖=1 𝑥𝑖 − 𝑦𝑖 . The name
relates to the distance a taxi has to drive in a rectangular street
grid
 Mahalanobis distance between random vector/points 𝒙 =
𝑥1 , ⋯ , 𝑥𝑛 and 𝒚 = 𝑦1 , ⋯ , 𝑦𝑛 is 𝒙 − 𝒚 𝑇 𝑺−1 𝒙 − 𝒚 ,
where 𝑺 is the covariance matrix
TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 9
 TOPSIS (contd…): Distance measure
 The Hamming distance between vector/points 𝒙 = 𝑥1 , ⋯ , 𝑥𝑛
and 𝒚 = 𝑦1 , ⋯ , 𝑦𝑛 is the number of positions at which the
corresponding values are different
 The 𝐿∞ norm between vector/points 𝒙 = 𝑥1 , ⋯ , 𝑥𝑛 and 𝒚 =
𝑦1 , ⋯ , 𝑦 is max 𝑦𝑖 − 𝑦𝑖
∀𝑖
 The 𝐿𝑝 norm between vector/points 𝒙 = 𝑥1 , ⋯ , 𝑥𝑛 and 𝒚 =
1
𝑦1 , ⋯ , 𝑦𝑛 is σ𝑛𝑖=1 𝑥𝑖 − 𝑦𝑖 𝑝 𝑝

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 10

 TOPSIS: Step # 01 (Construct the normalized
decision matrix)
𝑥11 ⋯ 𝑥1𝑛
 Assume the decision matrix, 𝑿 = ⋮ ⋱ ⋮
𝑥𝑚1 ⋯ 𝑥𝑚𝑛
𝑥𝑖𝑗
 Convert the entries in X into scaled normalized values, 𝑟𝑖𝑗 = , which has no
σ𝑚 2
𝑘=1 𝑥𝑘𝑗
dimension
𝑥11 𝑥1𝑛
⋯
𝑟11 ⋯ 𝑟1𝑛 σ𝑚 2
𝑘=1 𝑥𝑘1 σ𝑚 2
𝑘=1 𝑥𝑘𝑛

 Thus we get 𝑹 = ⋮ ⋱ ⋮ = ⋮ ⋱ ⋮
𝑟𝑚1 ⋯ 𝑟𝑚𝑛 𝑥𝑚1 𝑥𝑚𝑛
⋯
σ𝑚 2
𝑘=1 𝑥𝑘1 σ𝑚 2
𝑘=1 𝑥𝑘𝑛

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 11

 TOPSIS: Step # 01 (Construct the normalized
decision matrix) (contd..)
10 25 25 30 15
05 15 35 40 10
 Assume, 𝑿 =
15 25 40 45 10
20 30 30 35 05
𝑥𝑖𝑗
 Scale the values using normalization concept, i.e.,𝑟𝑖𝑗 = (you can use any other concept
σ𝑚 2
𝑘=1 𝑥𝑘𝑗
of utility also)
 𝑹= 10 625 625 900 225
100+25+225+400 625+225+625+900 625+1225+1600+900 900+1600+2025+1225 225+100+100+25
5 225 1225 1600 100
100+25+225+400 625+225+625+900 625+1225+1600+900 900+1600+2025+1225 225+100+100+25
15 625 1600 2025 100
100+25+225+400 625+225+625+900 625+1225+1600+900 900+1600+2025+1225 225+100+100+25
20 900 900 1225 25
100+25+225+400 625+225+625+900 625+1225+1600+900 900+1600+2025+1225 225+100+100+25

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 12

 TOPSIS: Step # 01 (Construct the normalized
decision matrix) (contd..)
𝑹 =
0.133333 0.263158 0.143678 0.156522 0.500000
0.033333 0.094737 0.281609 0.278261 0.222222
0.300000 0.263158 0.367816 0.352174 0.222222
0.533333 0.378947 0.206897 0.213043 0.055556
Check each column adds up to 1 as it should be

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 13

 TOPSIS: Step # 02 (Construct the weighted
normalized decision matrix)
 If the decision maker decides on the set of weights, depending on
𝑤1 ⋯ 0
his/her preference, then the weight, 𝑾 = ⋮ ⋱ ⋮ , such that σ𝑛𝑗=1 𝑤 =
0 ⋯ 𝑤𝑛
1
0.20 0.00 0.00 0.00 0.00
0.00 0.10 0.00 0.00 0.00
 Consider, 𝑾 = 0.00 0.00 0.15 0.00 0.00
0.00 0.00 0.00 0.25 0.00
0.00 0.00 0.00 0.00 0.30
 Calculate 𝑽 = 𝑹𝑾

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 14

 TOPSIS: Step # 02 (Construct the weighted
normalized decision matrix) (contd..)
0.20 0.00 0.00 0.00 0.00
0.133333 0.263158 0.143678 0.156522 0.500000
0.00 0.10 0.00 0.00 0.00
0.033333 0.094737 0.281609 0.278261 0.222222
 Thus 𝑽 = 𝑹𝑾 = × 0.00 0.00 0.15 0.00 0.00
0.300000 0.263158 0.367816 0.352174 0.222222
0.00 0.00 0.00 0.25 0.00
0.533333 0.378947 0.206897 0.213043 0.055556
0.00 0.00 0.00 0.00 0.30
𝑣11 𝑣21 𝑣31 𝑣41 𝑣51 0.026667 0.026316 0.021552 0.039130 0.150000
𝑣12 𝑣22 𝑣32 𝑣42 𝑣52 0.006667 0.009474 0.042241 0.069565 0.066667
 Hence 𝑽 = 𝑹𝑾 = 𝑣 𝑣 𝑣33 𝑣43 𝑣53 =
13 23 0.060000 0.026316 0.055172 0.088043 0.066667
𝑣14 𝑣24 𝑣34 𝑣44 𝑣54 0.106667 0.037895 0.031034 0.053261 0.016667

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 15

TOPSIS: Step # 03 (Determine the most positive-ideal, most
negative-ideal solutions)

Calculate 𝑽+ which is most positive ideal solution

Where 𝑽 + = max 𝑣𝑖𝑗 𝑗 ∈ 𝐽 , 𝑖 = 1, ⋯ , 𝑚 =
𝑖
+ + +
𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 16

TOPSIS: Step # 03 (Determine the most positive-ideal, most
negative-ideal solutions)

Calculate 𝑽− which is most negative ideal

solution
−
Where 𝑽 = min 𝑣𝑖𝑗 𝑗 ∈ 𝐽 , 𝑖 = 1, ⋯ , 𝑚 =
𝑖
𝑣1− , 𝑣2− , ⋯ , 𝑣𝑛−

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 17

 TOPSIS: Step # 04 (Calculate the distance based on most
positive-ideal, most negative-ideal solutions)
+
Calculate 𝑆𝑖 based on most positive ideal
solution
+ 𝑛 + 2
Where 𝑆𝑖 = σ𝑗=1 𝑣𝑖𝑗 − 𝑣𝑗 , 𝑖=
1, ⋯ , 𝑚

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 18

 TOPSIS: Step # 04 (Calculate the distance based on most
positive-ideal, most negative-ideal solutions)
−
Calculate 𝑆𝑖 based on most negative ideal
solution
− 𝑛 − 2
Where 𝑆𝑖 = σ𝑗=1 𝑣𝑖𝑗 − 𝑣𝑗 ,𝑖 =
1, ⋯ , 𝑚

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 19

 TOPSIS: Step # 05 (Calculate the relative proximity based
on ideal solution)
−
𝑆𝑖
Calculate 𝑐𝑖 = − + ,𝑖 =
𝑆𝑖 +𝑆𝑖
1, ⋯ , 𝑚

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 20

 TOPSIS: Example
 Consider the problem related to buying a house/apartment among four (04) choices, where the
decision to buy the house/apartment is based on eleven (11) different parameters/criterion which are
City
Price
Loan availability/conditions
Location
Number of rooms
Safety
Proximity to markets
Proximity to schools
Proximity to hospitals
Facilities available
Resale condition

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 21

 TOPSIS: Example (contd..)
45.5 40.0 66.7 09.2 1.8 0.0 1.3 2.3 1.6 4.8 3.1
50.0 33.3 83.3 08.5 1.5 1.3 0.0 2.4 2.4 3.3 1.6
𝑿 =
40.0 42.9 62.5 06.8 1.3 0.8 0.0 1.2 1.8 4.2 2.2
41.7 25.0 77.8 10.6 2.3 1.8 1.9 3.0 3.2 3.8 1.3
𝑥𝑖𝑗
 Use the normalization formulae as ,𝑟𝑖𝑗 =
σ𝑛 𝑥 2
𝑘=1 𝑖𝑘

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 22

 TOPSIS: Example (contd..)
𝑹=
0.5116 0.5561 0.4565 0.5180 0.5097 0.0000 0.5647 0.4961 0.3427 0.5908 0.7169
0.5622 0.4629 0.5701 0.4785 0.4248 0.5508 0.0000 0.5177 0.5140 0.4062 0.3700
0.4498 0.5964 0.4278 0.3828 0.3681 0.3390 0.0000 0.2589 0.3855 0.5169 0.5087
0.4689 0.3475 0.5325 0.5968 0.6513 0.7627 0.8253 0.6471 0.6854 0.4677 0.3006
 For example
42.9
 𝑟3,2 = =0.5964
40.02 +42.92 +62.52+6.82 +1.32 +0.82 + 0.02 +1.22+1.82 +4.22 +2.22
3.0
 𝑟4,8 = =0.6471
41.72 +252+77.82+10.62 +2.32 +1.82+1.92 +3.02 +3.22 +3.82 +1.32

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 23

 TOPSIS: Example (contd..)
• Consider 𝑾 =
(0.119, 0.106, 0.115, 0.090, 0.113, 0.061, 0.064, 0.113, 0.096, 0.065, 0.058),
where σ𝑛=11
𝑗=1 𝑤𝑗 = (0.119 + 0.106 + 0.115 + 0.090 + 0.113 + 0.061 +

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 24

 TOPSIS: Example (contd..)
0.0609 0.0589 0.0525 0.0466 0.0576 0.0000 0.0361 0.0545 0.0358 0.0385 0.0417
0.0669 0.0491 0.0656 0.0430 0.0480 0.0336 0.0000 0.0569 0.0493 0.0265 0.0215
 𝐕=
0.0535 0.0632 0.0493 0.0331 0.0416 0.0207 0.0000 0.0427 0.0370 0.0337 0.0295
0.0558 0.0368 0.0613 0.0586 0.0736 0.0446 0.0528 0.0758 0.0658 0.0304 0.0117
 For example 0.000
0.000
0.000
0.000
0.000
 v2,11 = 0.5622 0.4629 0.5701 0.4785 0.4248 0.5508 0.0000 0.5177 0.5140 0.4062 0.3700 × 0.000 =0.0215
0.000
0.000
0.000
0.000
0.058

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 25

 TOPSIS: Example (contd..)
Calculate
V+=(0.0669, 0.0632, 0.0656, 0.0586, 0.0736, 0.0446,
0.0528, 0.0758, 0.0658, 0.0385, 0.0417), where 𝐕 + =
max vij j ∈ J , i = 1, ⋯ , m = v1+ , v2+ , ⋯ , vn+
i
v1+ =max 0.0609,0.0669,0.0535,0.0558 = 0.0669
+
v2 =max 0.0589,0.0491,0.0632,0.0368 = 0.0632

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 26

 TOPSIS: Example (contd..)
Calculate
V-=(0.0535, 0.0368 , 0.0493, 0.0331, 0.0416, 0.0000,
0.0000, 0.0427, 0.0358, 0.0265, 0.0117), where 𝐕 − =
− − −
min 𝑣𝑖𝑗 𝑗 ∈ 𝐽 , 𝑖 = 1, ⋯ , 𝑚 = 𝑣1 , 𝑣2 , ⋯ , 𝑣𝑛
𝑖
v1− =min 0.0609,0.0669,0.0535,0.0558 = 0.0535
−
v2 =min 0.0589,0.0491,0.0632,0.0368 = 0.0368

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 27

 TOPSIS: Example (contd..)
Calculate the distance of each project to most positive
+ 2
ideal solution using 𝑆𝑖+ = σ𝑛𝑗=1 𝑣𝑖𝑗 − 𝑣𝑗 , thus
𝑆1+ = 0.1045
𝑆2+ = 0.1543
𝑆3+ = 0.0870
+
𝑆4 = 0.0425

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 28

 TOPSIS: Example (contd..)
Calculate the distance of each project to most negative
− 2
ideal solution using 𝑆𝑖− = σ𝑛𝑗=1 𝑣𝑖𝑗 − 𝑣𝑗 , thus
1
𝑆1 = 0.0589
−
𝑆2 = 0.0477
𝑆3− = 0.0387
−
𝑆4 = 0.0928

TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 29

 TOPSIS: Example (contd..)
 Calculate the relative proximity index of each alternative (which
is buying the house/apartment)
−
to the ideal solution according to
𝑆𝑖
formula 𝑐𝑖 = − + , thus
𝑆𝑖 +𝑆𝑖
0.0589
𝑐1 = = 0.3605
0.0589+0.1045
0.0477
𝑐2 = = 0.2361
0.0477+0.1543
0.0387
𝑐3 = = 0.3078
0.0387+0.0870
0.0928
𝑐3 = = 0.6853
0.0928+0.0425
TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 30
 TOPSIS: Example (contd..)
Thus the ranking is 𝑐4 ≻ 𝑐1 > 𝑐3 > 𝑐2
Hence Alternative # 04 is the best (position
# 01) choice followed by Alternative # 01
(position # 02), then by Alternative # 03
(position # 03) and finally followed by
Alternative # 02 (position # 04)
TOPSIS RNSengupta,IME Dept.,IIT Kanpur,INDIA 31