Scribd
Upload
242 views29 pages

7 Topsis Method

The document describes the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method for multi-criteria decision making. TOPSIS identifies the alternative that is closest to the ideal solution and farthest from the negative-ideal solution. The method involves 6 steps: 1) constructing a normalized decision matrix, 2) determining weighted normalized values, 3) identifying ideal and negative-ideal solutions, 4) calculating separation measures, 5) relative closeness scores, and 6) ranking alternatives based on closeness to the ideal solution. An example of choosing a new car using TOPSIS is provided to illustrate the steps.

Uploaded by

Allien Wang
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
242 views29 pages

7 Topsis Method

The document describes the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method for multi-criteria decision making. TOPSIS identifies the alternative that is closest to the ideal solution and farthest from the negative-ideal solution. The method involves 6 steps: 1) constructing a normalized decision matrix, 2) determining weighted normalized values, 3) identifying ideal and negative-ideal solutions, 4) calculating separation measures, 5) relative closeness scores, and 6) ranking alternatives based on closeness to the ideal solution. An example of choosing a new car using TOPSIS is provided to illustrate the steps.

Uploaded by

Allien Wang
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 29

DECISION MAKING WITH

MULTIPLE OBJECTIVES:
ELECTRE AND TOPSIS
METHODS
(PART B)

STQM6034 Decision and Game Analysis


TECHNIQUE FOR ORDER
PREFERENCE BY SIMILARITY
TO IDEAL SITUATION (TOPSIS)

STQM6034 Decision and Game Analysis


History

 Developed by Yoon and Hwang in 1980

 An alternative to ELECTRE method, and one of


the most widely accepted variants of
ELECTRE.
Method

 Considers three types of criteria/attributes

◦ Qualitative benefit criteria/attributes


◦ Quantitative benefit criteria/attributes
◦ Cost criteria/attributes
Basic Concept
the one which has the best level for all attributes considered

 The chosen decision alternative should be:

◦ as close (shortest distance) to the ideal


solution as possible, and
Artificial alternatives hypothesized

◦ as far (farthest distance) from the negative-


ideal solution as possible.

the one which has the worst level for all attributes considered
Assumption
1. Each criterion has a tendency of
monotonically increasing or decreasing
utility.

 Use Euclidean distance approach to evaluate


the relative closeness of the decision
alternatives to the ideal solution.

 Thus, the preference order of decision


alternatives can be derived by a series of
comparisons of these relative distances.
Decision matrix /
Inputs to method
 Given
◦ A set of m decision alternatives A1, A2, …, Am
◦ A set of n decision criteria/attributes C1, C2, …, Cn
◦ An m × n matrix X with xij performance measure of
the ith decision alternative in terms of the jth criterion
(i = 1,2,…,m and j = 1,2,…,n)
Criteria
Dec C1 C2 C3 … Cn
Alts
A1 x11 x12 x13 … x1n
X= A2 x21 x22 x23 … x2n
: : : : : :
Am xm1 xm2 xm3 … xmn
Decision matrix /
Inputs to method …continued

 Let
◦ J be the set of benefit attributes or criteria
(more is better)

◦ Let J’ be the set of negative attributes or


criteria (less is better)
STEPS

STQM6034 Decision and Game Analysis


Step 1: Construct the
Normalized Decision Matrix
 Convert the various criteria dimensions into
non-dimensional criteria.

 An element rij of the normalized decision


matrix R is calculated as:

xij
rij =
√ k=1 xkj
m 2
Step 2: Construct the Weighted
Normalized Decision Matrix
 A set of weights W = (w1, w2, …, wn) [where
total weights is 1] is determined by the
decision maker.

 Taking W and R together, the weighted


normalized decision matrix V is generated.

w1r11 w2r12 w3r13 … wnr1n


w1r21 w2r22 w3r23 … wnr2n
V= : : : : :
w1rm1 w2rm2 w3rm3 … wnrmn
Step 3a: Determining the Ideal
Solution
 The ideal decision alternative A* is defined as:

A* = { (max vij | jJ) , (min vij | jJ’), i= 1,2,3,…,m}


i i
= { v1*, v2*, …, vn* }

where:
J = { j = 1,2,3,…,n and j is associated with benefit
criteria}
J’ = { j = 1,2,3,…,n and j is associated with cost/loss
criteria}
Step 3b: Determining the
Negative-Ideal Solution
 The negative-ideal decision alternative A- is
defined as:

A- = { (min vij | jJ) , (max vij | jJ’), i= 1,2,3,…,m}


i i
= { v1-, v2-, …, vn- }

where:
J = { j = 1,2,3,…,n and j is associated with benefit
criteria}
J’ = { j = 1,2,3,…,n and j is associated with cost/loss
criteria}
Step 3: Some notes
 A* and A- are fictitious.

 However, it is reasonable to assume that:


◦ For benefit criteria, decision maker wants to
have a maximum value among the decision
alternatives
◦ For cost criteria, decision maker wants to have a
minimum value among the decision alternatives

 A* is the most preferable decision alternative.


 A- is the least preferable decision alternative.
Step 4: Calculate the Separation
Measure

 Need to measure the separation distances of each


decision alternative from the ideal solution and
negative-ideal solution.

 Apply the n-dimensional Euclidean distance


method
Step 4a: Calculate the
Separation Measure from Ideal
Solution
 The separation distances of each decision
alternative from the ideal solution is:

√
n

si* =  (vij – vj*)2


j=1

, for i = 1, 2, 3, …, m

Euclidean distance
Step 4b: Calculate the
Separation Measure from
Negative-Ideal Solution
 The separation distances of each decision
alternative from the negative-ideal solution is:

√
n

Si- =  (vij – vj-)2


j=1

, for i = 1, 2, 3, …, m

Euclidean distance
Step 5: Calculate the Relative
Closeness to the Ideal Solution
 The relative closeness of a decision alternative Ai
with respect to the ideal solution A* is defined as:

si-
Ci* =
si* + si-
where 0 ≤ Ci* ≤ 1, and i = 1, 2, 3, …, m

 Ci* = 1, if Ai = A*, and Ci* = 0, if Ai = A-


Step 6: Rank the Preference
Order

 The BEST (OPTIMAL) decision alternative is decided


according to the preference rank order Ci*.

 The BEST (OPTIMAL) decision alternative is the one


that has the shortest distance to the ideal solution.

 This also guarantees that the BEST (OPTIMAL)


decision alternative has the farthest distance from
the negative-ideal solution.
EXAMPLE

STQM6034 Decision and Game Analysis


Buying a new car
 Suppose the decision matrix of buying a new car is
given as below:

Style Reliability Fuel Cost


Economy
(0.1 0.4 0.3 0.2)
Car A 7 9 9 8
Car B 8 7 8 7
Car C 9 6 8 9
Car D 6 7 8 6

 Let J = {Style, Reliability, Fuel Economy}


 Let J’ = {Cost}
Step 1: Normalized Decision
Matrix R
 Refer to Excel file for workings.

Style Reliability Fuel Cost


Economy
(0.1 0.4 0.3 0.2)
Car A 0.462 0.614 0.545 0.528
Car B 0.528 0.477 0.484 0.462
Car C 0.593 0.409 0.484 0.593
Car D 0.396 0.477 0.484 0.396
Step 2: Weighted Normalized
Decision Matrix V
 Refer to Excel file for workings.

Style Reliability Fuel Cost


Economy
(0.1 0.4 0.3 0.2)
Car A 0.046 0.246 0.163 0.106
Car B 0.053 0.191 0.145 0.092
Car C 0.059 0.164 0.145 0.119
Car D 0.040 0.191 0.145 0.079
Step 3a: Determining the Ideal
Solution A*
 Refer to Excel file for workings.
 Let J = {Style, Reliability, Fuel Economy} Find max
 Let J’ = {Cost} Find min
Style Reliability Fuel Cost
Economy
(0.1 0.4 0.3 0.2)
Car A 0.046 0.246 0.163 0.106
Car B 0.053 0.191 0.145 0.092
Car C 0.059 0.164 0.145 0.119
Car D 0.040 0.191 0.145 0.079

 A* = {0.059, 0.246, 0.163, 0.079}


Step 3b: Determining the
Negative-Ideal Solution A -

 Refer to Excel file for workings.


 Let J = {Style, Reliability, Fuel Economy} Find min
 Let J’ = {Cost} Find max
Style Reliability Fuel Cost
Economy
(0.1 0.4 0.3 0.2)
Car A 0.046 0.246 0.163 0.106
Car B 0.053 0.191 0.145 0.092
Car C 0.059 0.164 0.145 0.119
Car D 0.040 0.191 0.145 0.079

 A- = {0.040, 0.164, 0.145, 0.119}


Step 4: Calculate the Separation
Measure from Ideal (Si*) and
Negative-Ideal Solutions (Si-)
 Refer to Excel file for workings.

Si* Si-

Car A 0.02949 Car A 0.08512


Car B 0.05936 Car B 0.04017
Car C 0.09270 Car C 0.01978
Car D 0.06081 Car D 0.04806
Step 5: Calculate the Relative
Closeness to the Ideal Solution
(Ci*) and Step 6
 Refer to Excel file for workings.

Ci*

Car A 0.743 MAX = BEST


Car B 0.404
Car C 0.092
Car D 0.441
SELF STUDY

STQM6034 Decision and Game Analysis


Check out the variants and
hybrids of TOPSIS

 Find out the other variants and hybrids of TOPSIS.

 For each variant and hybrid, check and see how


they are different/improved/modified.

 Also look at the suitable conditions or decision


making problem that can benefit from these
variants and hybrids.

You might also like