Vous êtes sur la page 1sur 9

Neutrosophic Sets and Systems, Vol.

17, 2017 20

University of New Mexico

An extended TOPSIS for multi-attribute decision making


problems with neutrosophic cubic information
1 2 3 4
Surapati Pramanik , Partha Pratim Dey , Bibhas C. Giri , and Florentin Smarandache
1
Department of Mathematics, Nandalal Ghosh B.T. College, Panpur, P.O.-Narayanpur, District - North 24 Parganas, Pin Code-743126, West Bengal,
India. E-mail: sura_pati@yahoo.co.in
2
Department of Mathematics, Jadavpur University, Kolkata-700032, West Bengal, India. E-mail: parsur.fuzz@gmail.com
3
Department of Mathematics, Jadavpur University, Kolkata-700032, West Bengal, India. E-mail: bcgiri.ju.math@gmail.com
4
University of New Mexico. Mathematics & Science Department, 705 Gurley Ave., Gallup, NM 87301, USA. Email: fsmarandache@gmail.com

Abstract. The paper proposes a new technique for deal- predefined attributes are presented in terms of neutro-
ing with multi-attribute decision making problems sophic cubic information where weights of the attributes
through an extended TOPSIS method under neutrosophic are completely unknown. In the selection process, neu-
cubic environment. Neutrosophic cubic set is the general- trosophic cubic positive and negative ideal solutions have
ized form of cubic set and is the hybridization of a neu- been defined. An extended TOPSIS method is then pro-
trosophic set with an interval neutrosophic set. In this posed for ranking the alternatives and finally choosing
study, we have defined some operation rules for neutro- the best one. Lastly, an illustrative example is solved to
sophic cubic sets and proposed the Euclidean distance demonstrate the decision making procedure and effec-
between neutrosophic cubic sets. In the decision making tiveness of the developed approach.
situation, the rating of alternatives with respect to some

Keywords: TOPSIS; neutrosophic sets; interval neutrosophic set; neutrosophic cubic sets; multi-attribute decision making.

1 Introduction obtained by maximizing deviation method. Ye [8]


proposed an extended TOPSIS method for solving MADM
Smarandache [1] proposed neutrosophic set (NS) that
problems under interval neutrosophic uncertain linguistic
assumes values from real standard or non-standard subsets
variables. Biswas et al. [9] studied TOPSIS method for
of] -0, 1+[. Wang et al. [2] defined single valued
solving multi-attribute group decision making problems
neutrosophic set (SVNS) that assumes values from the unit
with single-valued neutrosophic information where
interval [0, 1]. Wang et al. [3] also extended the concept of
weighted averaging operator is employed to aggregate the
NS to interval neutrosophic set (INS) and presented the
individual decision makers opinion into group opinion.
set-theoretic operators and different properties of INSs.
In 2016, Ali et al. [10] proposed the notion of neutrosophic
Multi-attribute decision making (MADM) problems with
cubic set (NCS) by extending the concept of cubic set to
neutrosophic information caught much attention in recent
neutrosophic cubic set. Ali et al. [10] also defined internal
years due to the fact that the incomplete, indeterminate and
neutrosophic cubic set (INCS) and external neutrosophic
inconsistent information about alternatives with regard to
predefined attributes are easily described under cubic set (ENCS), 1 3 -INCS ( 2 3 -ENCS), 2 3 -INCS
neutrosophic setting. In interval neutrosophic environment,
( 1 3 -ENCS) and also proposed some relevant properties.
Chi and Liu [4] at first established an extended technique
for order preference by similarity to ideal solution In the same study, Ali et al. [10] proposed Hamming
(TOPSIS) method [5] for solving MADM problems with distance between two NCSs and developed a decision
interval neutrosophic information to get the most making technique via similarity measures of two NCSs in
preferable alternative. ahin, and Yiider [6] discussed pattern recognition problems. Jun et al. [11] studied the
TOPSIS method for multi-criteria decision making notions of truth-internal (indeterminacy-internal, falsity-
(MCDM) problems with single neutrosophic values for internal) neutrosophic cubic sets and truth-external
supplier selection problem. Zhang and Wu [7] developed (indeterminacy-external, falsity-external) neutrosophic
an extended TOPSIS for single valued neutrosophic cubic and investigated related properties. Pramanik et al.
MCDM problems where the incomplete weights are [12] defined similarity measure for neutrosophic cubic sets

Surapati Pramanik, Partha Pratim Dey, Bibhas C. Giri, Florentin Smarandache. An extended TOPSIS for multi-attribute
decision making problems with neutrosophic cubic information
Neutrosophic Sets and Systems, Vol. 17, 2017 21

and proved its basic properties. In the same study, where and respectively represent an interval valued
Pramanik et al. [12] developed multi criteria group deci- fuzzy set and a fuzzy set. A cubic set is denoted by =
sion making method with linguistic variables in neutrosophic < , >.
cubic set environment.
Definition: 4
In this paper, we develop a new approach for MADM Internal cubic sets [15]: A cubic set = < , > in U is
problems with neutrosophic cubic assessments by using said to be internal cubic set (ICS) if
TOPSIS method where weights of the attributes are un-
known to the decision maker (DM). We define a few oper- (x) (x) (x) for all x U.
-

ations on NCSs and propose the Euclidean distance be-


tween two NCSs. We define accumulated arithmetic opera- Definition: 5
tor (AAO) to convert neutrosophic cubic values (NCVs) to External cubic sets [15]: A cubic set = < , > in U is
single neutrosophic values (SNVs). We also define neutro- called external cubic set (ECS) if (x) ( - (x) , (x) )
sophic cubic positive ideal solution (NCPIS) and neutro-
for all x U.
sophic cubic negative ideal solution (NCNIS) in this study.
The rest of the paper is organized in the following way.
Section 2 recalls some basic definitions which are useful Definition: 6
for the construction of the paper. Subsection 2.1 provides Consider 1 = < 1 , 1 > and 2 = < 2 , 2 > be two cubic
several operational rules of NCSs. Section 3 is devoted to sets in U, then we have the following relations [15].
present an extended TOPSIS method for MADM problems 1. (Equality) 1 = 2 if and only if 1 = 2
in neutrosophic cubic set environment. In Section 4, we
and 1 = 2 .
solve an illustrative example to demonstrate the applicabil-
ity and effectiveness of the proposed approach. Finally, the
last Section presents concluding remarks and future scope 2. (P-order) 1 P 2 if and only
research. if 1 2 and 1 2 .

2 The basic definitions 3. (R-order) 1 R 2 if and only


Definition: 1 if 1 2 and 1 2 .
Fuzzy sets [13]: Consider U be a universe. A fuzzy set
over U is defined as follows: Definition: 7
= { x, ( x) x U} Neutrosophic set [1]: Consider U be a space of objects,
where (x) : U [0, 1] is termed as the membership then a neutrosophic set (NS) on U is defined as follows:
function of and (x) represents the degree of mem- = {x, ( x), ( x), ( x) xU}
- +
bership to which x . where ( x), ( x), ( x) :U ] 0, 1 [ define respectively the
degrees of truth-membership, indeterminacy-membership,
Definition: 2 and falsity-membership of an element xU to the
Interval valued fuzzy sets [14]: An interval-valued fuzzy set with -0 sup (x) + sup (x) + sup (x) 3+.
set (IVFS) over U is represented as follows:
Definition: 8
= { x, - ( x), ( x) x U}
Interval neutrosophic sets [9]: An INS in the
where - ( x), ( x) denote the lower and upper degrees universal space U is defined as follows:
of membership of the element x U to the set with = {x, [ - ( x), ( x)],[ - ( x), ( x)],[ - ( x), ( x)]
0 (x) + ( x ) 1.
-
x U}
where, (x) , (x ) , (x ) are the truth-membership
Definition: 3 function, indeterminacy-membership function, and falsity-
Cubic sets [15]: A cubic set in a non-empty set U is a
membership function, respectively with (x) , (x ) ,
structure defined as follows:
= { x, ( x ), ( x) x U} (x ) [0, 1] for each point x U and 0 sup (x) +
sup (x ) + sup (x ) 3.

Surapati Pramanik, Partha Pratim Dey, Bibhas C. Giri, Florentin Smarandache, An extended TOPSIS for multi-attribute
decision making problems with neutrosophic cubic information
22 Neutrosophic Sets and Systems, Vol. 17, 2017

Definition: 9 Example 3. Consider = < , > NCSU, if ( x ) = <


Neutrosophic cubic sets [15] [0.65, 0.8], [0.1, 0.25], [0.2, 0.4] > and (x) = < 0.85, 0.3,
A neutrosophic cubic set (NCS) in a universe U is
presented in the following form: 0.1 > for all x U, then = < , > is an ENCS.
= { x, ( x ), ( x) x U}
Theorem 1. [10]
where and are respectively an interval neutrosophic set
Consider = < , > NCSU, which is not an ENCS,
and a neutrosophic set in U.
However, NSs take the values from] -0, 1+[ and single- then there exists xU such that
valued neutrosophic set defined by Wang et al. [2] is - (x) (x) (x) ; - (x ) (x) (x ) ; or
appropriate for dealing with real world problems. - (x) (x) (x ) .
Therefore, the definition of NCS should be modified in
order to solve practical decision making purposes. Hence,
a neutrosophic cubic set (NCS) in U is defined as Definition: 12
2 -INCS( 1 -ENCS) [10]: Consider = < , > NCSU,
follows:
3 3
= { x, ( x ), ( x) x U}
Here, and are respectively an INS and a SVNS in U
if (x) (x) (x ) ;
-


- (x) (x) (x) ;
where 0 (x) + (x) + (x) 3 for each point x U. and (x) ( - (x) , (x ) ) or - (x) (x) (x) ;
Generally, a NCS is denoted by = < , > and sets of - (x) (x) (x ) and (x) ( - (x ) , (x ) ) or
all NCS over U will be represented by NCSU.
- (x ) (x) (x ) ; and - (x) (x) (x )
Example 1. Assume that U = {u1, u2, u3, u4} be a universal and (x) ( - (x) , (x) ) for all x U, then = < ,
set. An INS A in U is defined as
= {< [0.15, 0.3], [0.25, 0.4], [0.6, 0.75] >/ u1 + < [0.25, > is said to be an 2 -INCS or 1 -ENCS.
3 3
0.35], [0.35, 0.45], [0.4, 0.65] >/ u2 + < [0.35, 0.5], [0.25, Example 4. Consider = < , > NCSU, if ( x ) = <
0.35], [0.55, 0.85] >/ u3 + < [0.7, 0.8], [0.15, 0.45], [0.2,
0.3] >/ u4} [0. 5, 0.7], [0.1, 0.2], [0.2, 0.45] > and (x) = < 0.65, 0.3,
and a SVNS in U defined by 0.4 > for all x U, then = < , > is an 2 -INCS or 1 -
= {< 0.35, 0.3, 0.15 >/ u1, < 0.5, 0.1, 0.4 >/ u2, < 0.25, 3 3
ENCS.
0.03, 0.35 >/ u3, < 0.85, 0.1, 0.15 >/ u4}
Then = < A, > is represented as a NCS in U.
Definition: 13
1 -INCS ( 2 -ENCS) [10]: Consider = < , >
Definition: 10 3 3
Internal neutrosophic cubic set [10]: Consider = < ,
NCSU, if - (x) (x) (x) ; (x) ( - (x ) , (x ) );
> NCSU, if - (x) (x) (x) ;
and (x) ( - (x) , (x ) ) or - (x) (x) (x ) ;
- (x ) (x) (x ) ; and - (x) (x) (x ) for
all x U, then is said to be an internal neutrosophic (x) ( - (x) , (x) ) and (x) ( - (x ) , (x ) ) or
cubic set (INCS). - (x ) (x) (x ) ; (x) ( - (x) , (x) ) and

Example 2. Consider = < , > NCSU, if ( x ) = < (x) ( - (x) , (x ) ) for all x U, then = < , > is
[0.65, 0.8], [0.1, 0.25], [0.2, 0.4] > and (x) = < 0.7, 0.2, said to be an 1 -INCS or 2 -ENCS.
3 3
0.3 > for all x U, then = < , > is an INCS.
Example 5. Consider = < , > NCSU, if (x) = <
Definition: 11 [0. 5, 0.8], [0.15, 0.25], [0.2, 0.35 ] > and (x) = < 0.55,
External neutrosophic cubic set [10]: Consider = < , 0.4, 0.5 > for all x U, then = < , > is an 1 -INCS
> NCSU, if (x) ( - (x) , (x) ); 3
or 2 -ENCS.
(x) ( - (x ) , (x ) ); and (x) ( - (x) , (x ) ) for 3
all x U, then = < , > is said to be an external Definition: 14 [10]
neutrosophic cubic set (ENCS). Consider 1 = < 1 , 1 > and 2 = < 2 , 2 > be two
NCSs in U, then

________________________________________________________________________________________________
Surapati Pramanik, Partha Pratim Dey, Bibhas C. Giri, Florentin Smarandache. An extended TOPSIS for multi-attribute
decision making problems with neutrosophic cubic information
Neutrosophic Sets and Systems, Vol. 17, 2017 23

1. (Equality) 1 = 2 if and only if 1 = 2 4. Consider p1 = < ([ - , ], [ , ], [ - , ]),


1 1 1 1 1 1

and 1 = 2 . ( 1 , 1 , 1 ) > be a NCV and be an arbitrary positive


~ real number, then p1and p1 are defined as follows:
2. (P-order) 1 P 2 if and only if 1 2

and 1 2 .
(i) p1= < ([1- (1- - ) ,1- (1- ) ], 1 1

3. (R-order) 1 R 2 if and only [( ) , ( ) ],[( ) , ( ) ]), (1- (1-



1 1 1 1
~
if 1 2 and 1 2 .
1 ) , ( 1 ) , ( 1 ) ) >;

For convenience, p = < ([ - , ], [ , ], [ - , ]), 1 1 1 (ii) p1 = < ([( - ) , ( ) ], [1- (1- ) ,1- (1-
1 1 1

( , , ) > is said to represent neutrosophic cubic value 1 ) ], [1- (1- 1 ) ,1- (1- 1 ) ]),
(NCV)
(( 1 ) ,1-(1- 1 ) ,1- (1- 1 ) ) >.
Definition: 15
Complement [10]: Consider = < , > be an NCS, Definition: 16 [10]
then the complement of = < , > is given by Consider 1 = < 1 , 1 > and 2 = < 2 , 2 > be two
~ NCSs in U, then the Hamming distance between 1 and
C = { x, C ( x ), C ( x) x U}.
2 is defined as follows:
2.1 Several operational rules of NCVs 1 n
DH ( 1 , 2 ) = (| ( x ) - ( x ) | + | ( x ) -
- - -
1 i 2 i 1 i
9 n i 1
Consider p1 = < ([ - , ], [ , ], [ - , ]),
1 1 1 1 1 1 2- ( xi ) | + | 1- ( xi ) - 2- ( xi ) | + | 1 ( xi ) - 2 ( xi ) | +
( 1 , 1 , 1 ) > and p2 = < ([ - , ], [ , ], 2 2 2 2 | 1 ( xi ) - 2 ( xi ) | + | 1 ( xi ) - 2 ( xi ) | + | 1 ( xi ) - 2 ( xi ) |

[ , ]), ( 2 , 2 , 2 ) > be two NCVs in U, then the
-
2 2 + | 1 ( xi ) - 2 ( xi ) | + | 1 ( xi ) - 2 ( xi ) |).
operational rules are presented as follows:
1. The complement [10] of p1 is p1C = < ([ - , ], [1- 1 1
Example 7: Suppose that 1 = < 1 , 1 > = < ([0.6, 0.75],

,1- ], [ , ]), ( 1 , 1- 1 , 1 ) >.
1 1
-
1 1
[0.15, 0.25], [0.25, 0.45]), (0.8, 0.35, 0.15) > and 2 =
< 2 , 2 > = < ([0.45, 0.7], [0.1, 0.2], [0.05, 0.2]), (0.3,
2. The summation between p1 and p2 is defined as 0.15, 0.45) > be two NCSs in U, then DH ( 1 , 2 ) =
follows:
0.1944.
p1 p2 = < ([ - + - - - - 2 , 1 + -
1 2 1 2 Definition: 17

], [
,
], Consider 1 = < 1 , 1 > and 2 = < 2 , 2 > be two
1 2 1 2 1 2
NCSs in U, then the Euclidean distance between 1 and
[ -
1 , -
2 1 ]), ( 1 + 2 - 1 2 , 1 2 , 1 2 )
2
2 is defined as given below.
>.
3. The multiplication between p1 and p2 is defined as DE ( 1 , 2 ) =
follows: ( 1- ( xi ) - 2- ( xi ))2 ( 1 ( xi ) - 2 ( xi ))2 ( 1- ( xi ) - 2- ( xi ))2

1 n
( ( x ) - ( x )) ( ( x ) - ( x )) ( ( x ) - ( x ))
2 - - 2 2
p1 p2= < ([ - - , ], [ + - 9n i 1
1 i 2 i 1 i 2 i 1 i 2 i

(1 ( xi ) - 2 ( xi ))2 ( 1 ( xi ) - 2 ( xi ))2 (1 ( xi ) - 2 ( xi ))2
1 2 1 2 1 2



,
+
-

], [ -
+ -
-
1 2 1 2 1 2 1 2

-
1 , + -
-
2 1 2 1 ]), ( 1 2 , 1 + 2 -
2 with the condition 0 DE ( 1 , 2 ) 1.
1 2 , 1 + 2 - 1 2 ) >.
Example 8: Suppose that 1 = < 1 , 1 > = < ([0.4, 0.5],
[0.1, 0.2], [0.25, 0.5]), (0.4, 0.3, 0.25) > and 2 = < 2 ,
2 > = < ([0.5, 0.9], [0.15, 0.3], [0.05, 0.1]), (0.7, 0.1,

Surapati Pramanik, Partha Pratim Dey, Bibhas C. Giri, Florentin Smarandache. An extended TOPSIS for multi-attribute
decision making problems with neutrosophic cu
cubic information
24 Neutrosophic Sets and Systems, Vol. 17, 2017

0.15) > be two NCSs in U, then DE ( 1 , 2 ) = 0.2409. Here a ijc denotes the complement of aij.

3 An extended TOPSIS method for MADM prob- Step 2. Identify the weights of the attributes
lems under neutrosophic cubic set environment To determine the unknown weight of attribute in the
In this Section, we introduce a new extended TOPSIS decision making situation is a difficult task for DM.
method to handle MADM problems involving Generally, weights of the attributes are dissimilar and they
neutrosophic cubic information. Consider B = {B1, B2, , play a decisive role in finding the ranking order of the
Bm}, (m 2) be a discrete set of m feasible alternatives and alternatives. Pramanik and Mondal [16] defined arithmetic
C = {C1, C2, , Cn}, (n 2) be a set of attributes. Also, let averaging operator (AAO) in order to transform interval
w = (w1, w2, , wn)T be the unknown weight vector of the neutrosophic numbers to SVNNs. Based on the concept of
Pramanik and Mondal [16], we define AAO to transform
attributes with 0 wj 1 such that
n
w j = 1. Suppose that NCVs to SNVs as follows:
j 1

the rating of alternative Bi (i = 1, 2, , m) with respect to NCij < ij , ij , > =


ij

the attribute Cj (j = 1, 2, , n) is described by aij where aij


- ij - ij - ij
= < ([ -ij , ij ], [ ij , ij ], [ - , ]), ( ij , ij , ij ) >.
ij ij NCij
ij ij
, ij ij
, ij ij

3 3 3
The proposed approach for ranking the alternatives under
neutrosophic cubic environment is shown using the In this paper, we utilize information entropy method to
following steps: find the weights of the attributes wj where weihgts of the
attributes are unequal and fully unknown to the DM.
Step 1. Construction and standardization of decision Majumdar and Samanta [17] investigated some similarity
matrix with neutrosophic cubic information measures and entropy measures for SVNSs where entropy
is used to measure uncertain information. Here, we take the
Consider the rating of alternative Bi (i = 1, 2, , m) with following notations:
respect to the predefined attribute Cj, (j = 1, 2, , n) be - ij - ij
presented by the decision maker in the neutrosophic cubic T (xi) =
ij
, I (xi) =
ij
,
ij ij

decision matrix ( See eqn. 1).


P
3 P
3
a11 a12 ... a1n - ij

F (xi) =
ij ij

a 21 a 22 ... a 2 n P
3
aij = . . ... . (1)
mn
. . ... . Then we can write P = TP ( xi ), I P ( xi ), FP ( xi ) .
a a ... a The entropy value is given as follows:
m1 m 2 mn
1 m
Ei ( P ) = 1 - (T ( xi ) F ( xi )) I ( xi ) I c ( xi )
In general, there are two types of attributes appear in the n i 1 P P P P

decision making circumstances namely (i) benefit type which has the following properties:
attributes J1, where the more attribute value denotes (i). Ei ( P ) = 0 if P is a crisp set and I ( xi ) = 0,
better alternative (ii) cost type attributes J2, where the P

less attribute value denotes better alternative. We F ( xi ) = 0 x E.


P

standardize the above decision matrix aij in order to (ii). Ei ( P ) = 0 if TP ( xi ), I P ( xi ), FP ( xi ) = < T ( xi ) ,


mn P

remove the influence of diverse physical dimensions to 0.5, F ( xi ) >, x E.


P
decision results. (iii). Ei ( P ) Ei ( Q ) if P is more uncertain
Consider sij be the standardize decision matrix,
mn than Q i.e.
where T ( xi ) + F ( x i ) TQ ( xi ) + FQ ( xi )
s ij = < ([ - , ], [ , ], [ - , ]),
ij ij
P P

and I P ( xi ) I c P ( xi ) I Q ( xi ) I c Q ( xi ) .
ij ij ij ij

( ij , ij , ij )>,
where (iv). Ei ( P ) = Ei ( Pc ), x E.
s ij = aij, if the attribute j is benefit type; Consequently, the entropy value Evj of the j-th attribute
can be calculated as as follows:.
s ij = a ijc , if the attribute j is cost type.

________________________________________________________________________________________________
Surapati Pramanik, Partha Pratim Dey, Bibhas C. Giri, Florentin Smarandache. An extended TOPSIS for multi-attribute
decision making problems with neutrosophic cubic information
Neutrosophic Sets and Systems, Vol. 17, 2017 25

Evj = 1 -
1 m
(T ( x ) F ( x )) I ( x ) I ( x ) , i = 1, 2, ,
C ( - )U = {( max { -ij }| jJ1), ( min { -ij }| jJ2)},
j
ij i ij i ij i ij i i i
n i 1
m; j = 1, 2,, n. ( )U = {( max { ij }| jJ1), ( min { ij }| jJ2)},

j
i
We observe that 0 Evj 1. Based on Hwang and Yoon i

[18] and Wang and Zhang [19] the entropy weight of the ( )U = {( min { }| jJ1), ( max { }| jJ2)},
j ij ij
i i
j-th attribute is defined as follows:
1 Ev j (
j ) U
={( min {
ij }| jJ1), ( max {
ij }| jJ2)},
with 0 wj 1 and w j = 1.
n i i
wj = n
(1 Ev j )
j1
( - )U= {( min { - }| jJ1), (
j ij
max { - }| jJ2)},
ij
i i
j1

( )U ={( min {

}| jJ1), ( max { }| jJ2)},
j ij ij
i i
Step 3. Formulation of weighted decision matrix U
The weighted decision matrix is obtained by multiplying ( j ) = {( max { ij }| jJ1), ( min { ij }| jJ2)},
i i
weights of the attributes (wj) and the standardized decision
( j )U = {( min { ij }| jJ1), ( max { ij }| jJ2)},
matrix sij . Therefore, the weighted neutrosophic cubic i i
mn

decision matrix z ij is obtained as: ( j ) = {( min { ij }| jJ1), ( max { ij }| jJ2)};


U
i i
mn
L
w1s11 w2 s12 ... wn s1n and z j is defined as given below

w s w2 s22 ... wn s2 n z Lj = < [( -ij )L, ( ij )L], [( ij )L, ( ij )L], [( - )L,
= w j aij
ij
z ij = 1 21
mn mn
. . ... . ( )L], (( ij )L, ( ij )L, ( ij )L)>
ij
w s
1 m1 w2 sm 2 ... wn smn where ( - )L = {( min { -ij }| jJ1), (
j
max { - }| ij
i i
z11 z12 ... z1n
jJ2)}, ( )L = {( min { ij }| jJ1), ( max { ij }|
j

z 21 z 22 ... z 2 n i i

= . . ... . }| jJ1), ( min { }|


jJ2)}, ( ) = {( max { L
ij
j ij
i i
. . ... .
jJ2)}, ( )L ={( max { }| jJ1), ( min { }|
z z m 2 ... z mn j ij
i ij
m1 i

where jJ2)}, ( ) = {( -
j
L
max { }| jJ1), ( min { }|
-
ij
-
ij
i i
zij = < ([ -ij , ij ], [ ij , ij ], [ - , ]), ( ij , ij , ij ) >
ij ij jJ2)}, ( j )L ={( max { ij }| jJ1), ( min { }| ij
i i
= < ([1- (1- - ij ) ,1- (1- ij ) ],
wj wj
L
jJ2)}, ( j ) = {( min { ij }| jJ1), ( max { ij }| jJ2)},
i i
[( ij ) j , ( ij ) j ],[( ) , ( ij ) j ]), (1- (1- ij ) ,
w w w j w w j

ij
( j )L = {( max { ij }| jJ1), ( min { ij }| jJ2)}, ( j )L =
( ij ) , ( ij ) ) >
w j w j i i

{( max { ij }| jJ1), ( min { ij }| jJ2)}.


i i
Step 4. Selection of neutrosophic cubic positive ideal
solution (NCPIS) and neutrosophic cubic negative ideal Step 5. Calculate the distance measure of alternatives
solution (NCNIS) from NCPIS and NCNIS
Consider zU = ( z1U , z U2 , , z Un ) and zL = ( z1L , z 2L , , z nL ) The Euclidean distance measure of each alternative Bi, i =
1, 2, , m from NCPIS can be defined as follows:
be the NCPIS and NCNIS respectively, then z Uj is defined
DEi =
as follows:
z Uj = < ([( - )U, ( )U], [( j )U, ( j )U], [( - )U, ( - -(- ) U ) 2 ( -( )U ) 2 ( -( )U ) 2
j j j
ij j ij j ij j

1 n U 2
( )U]), (( j )U, ( j )U, ( j )U) > ( -(
) ) ( - ( ) ) ( -( ) )
- - U 2 U 2

9n j1
j ij j ij j ij j

where ( -( )U ) 2 ( -( ) U ) 2 ( -( ) U ) 2
ij j ij j ij j

Surapati Pramanik, Partha Pratim Dey, Bibhas C. Giri, Florentin Smarandache. An extended TOPSIS for multi-attribute
decision making problems with neutrosophic cu
cubic information
26 Neutrosophic Sets and Systems, Vol. 17, 2017

Similarly, the Euclidean distance measure of each Table 1. Neutrosophic cubic decision matrix
alternative Bi, i = 1, 2, , m from NCNIS can be written as C1 C2
follows: B1 ([0.3, 0.4],[0.1, 0.2],[0.2, 0.35]), ([0.6, 0.7],[0.05, 0.1],[0.2, 0.3]),
D E = (0.3, 0.4, 0.25) (0.5, 0.1, 0.25)
i
B2 ([0.8, 0.9],[0.1, 0.2],[0.15, 0.3]), ([0.3, 0.5],[0.1, 0.4],[0.3, 0.5]),
( - -( - ) L ) 2 ( -( ) L ) 2 ( -( ) L ) 2 (0.7, 0.15, 0.3) (0.4, 0.3, 0.2)
ij j ij j ij j
1 n L 2 - -( - ) L ) 2 ( -( ) L ) 2 . B3 ([0.6, 0.7],[0.2, 0.4],[0.25, 0.4]), ([0.2, 0.35],[0.1, 0.25],[0.2, 0.3]),
( -( ) ) (
9n j 1

(0.5, 0.3, 0.3) (0.3, 0.3, 0.4)
ij j ij j ij j

( -( ) L ) 2 ( -( ) L ) 2 ( -( ) L ) 2
ij j ij j ij j

Step 6. Evaluate the relative closeness co-efficient to the C3 C4


neutrosophic cubic ideal solution B1 ([0.5, 0.6],[0.2, 0.4],[0.1, 0.3]), ([0.4, 0.6],[0.1, 0.25],[0.1, 0.3]),
The relative closeness co-efficient RCCi* of each Bi, i = 1, (0.5, 0.3, 0.4) (0.5, 0.2, 0.4)
B2 ([0.4, 0.5],[0.2, 0.35],[0.05, 0.2]), ([0.2, 0.3],[0.2, 0.35],[0.1, 0.25]),
2, , m with respect to NCPIS z Uj , j = 1, 2, , n is (0.35, 0.1, 0.1) (0.4, 0.1, 0.1)
defined as follows: B3 ([0.4, 0.7],[0.1, 0.3],[0.15, 0.25]), ([0.5, 0.7],[0.1, 0.2],[0.2, 0.25]),
DE (0.5, 0.2, 0.2) (0.3, 0.1, 0.2)
RCCi* = i , i = 1, 2, , m.
DEi DEi
Step 2. Standardize the decision matrix.
The standardized decision matrix sij is constructed as
Step 7. Rank the alternatives 34
We obtain the ranking order of the alternatives based on follows (see Table 2):
the RCCi* . The bigger value of RCCi* reflects the better
alternative. Table 2. The standardized neutrosophic cubic decision
matrix
4. Numerical example C1 C2
B1 ([0.2, 0.35],[0.8, 0.9,[0.3, 0.4]), ([0.2, 0.3],[0.9, 0.95],[0.6, 0.7]),
In this section, we consider an example of neutrosophic (0.25, 0.6, 0.3) (0.25, 0.9, 0.5)
cubic MADM, adapted from Mondal and Pramanik [20] to B2 ([0.15, 0.3],[0.8, 0.9],[0.8, 0.9]), ([0.3, 0.5],[0.6, 0.9],[0.3, 0.5]),
demonstrate the applicability and the effectiveness of the (0.3, 0.85, 0.7) (0.2, 0.7, 0.4)
proposed extended TOPSIS method. B3 ([0.25, 0.4],[0.6, 0.8],[0.6, 0.7]), ([0.2, 0.3],[0.75, 0.9],[0.2, 0.35]),
Consider a legal guardian desires to select an appropriate (0.3, 0.7, 0.5) (0.4, 0.7, 0.3)
school for his/ her child for basic education [20]. Suppose
there are three possible alternatives for his/ her child: C3 C4
(1) B1, a Christian missionary school B1 ([0.5, 0.6],[0.2, 0.4],[0.1, 0.3]), ([0.4, 0.6],[0.1, 0.25],[0.1, 0.3]),
(2) B2, a Basic English medium school (0.5, 0.3, 0.4) (0.5, 0.2, 0.4)
(3) B3, a Bengali medium kindergarten. B2 ([0.4, 0.5],[0.2, 0.35],[0.05, 0.2]), ([0.2, 0.3],[0.2, 0.35],[0.1, 0.25]),
He/ She must take a decision based on the following four (0.35, 0.1, 0.1) (0.4, 0.1, 0.1)
attributes: B3 ([0.4, 0.7],[0.1, 0.3],[0.15, 0.25]), ([0.5, 0.7],[0.1, 0.2],[0.2, 0.25]),
(1) C1 is the distance and transport, (0.5, 0.2, 0.2) (0.3, 0.1, 0.2)
(2) C2 is the cost,
(3) C3 is the staff and curriculum, and
Step 3. Using AAO, we transform NCVs into SNVs. We
(4) C4 is the administrative and other facilities
calculate entropy value Ej of the j-th attribute as follows:
Here C1 and C2 are cost type attributes; while C3 and C4 are
Ev1 = 0.644, Ev2 = 0.655, Ev3 = 0.734, Ev4 = 0.663.
benefit type attributes. Suppose the weights of the four
The weight vector of the four attributes are obtained as:
attributes are unknown. Using the the following steps, we
w1 = 0.2730, w2 = 0.2646, w3 = 0.2040, w4 = 0.2584.
solve the problem.
Step 4. After identifying the weight of the attribute (wj),
Step 1. The rating of the alternative Bi, i = 1, 2, 3 with
we multiply the standardized decision matrix with wj, j = 1,
respect to the alternative Cj, j = 1, 2, 3, 4 is represented by
neutrosophic cubic assessments. The decision matrix 2, , n to obtain the weighted decision matrix z ij (see
34
aij is shown in Table 1. Table 3).
34

________________________________________________________________________________________________
Surapati Pramanik, Partha Pratim Dey, Bibhas C. Giri, Florentin Smarandache. An extended TOPSIS for multi-attribute
decision making problems with neutrosophic cubic information
Neutrosophic Sets and Systems, Vol. 17, 2017 27

Table 3. The weighted neutrosophic cubic decision matrix Step 7. The relative closeness co-efficient RCCi* , i = 1, 2,
C1 C2
([0.059, 0.110], [0.941, 0.972, [0.720, 0.779]),
3 is obtained as follows:
B1 ([0.057, 0.090,[0.972, 0.986],[0.874, 0.91]),
(0.075, 0.87, 0.72) (0.073, 0.972, 0.832) RCC1* = 0.3640, RCC2* = 0.4622, RCC3* = 0.3802.
B2 ([0.043, 0.093],[0.941, 0.972],[0.941, 0.972]), ([0.09, 0.168],[0.874, 0.972],[0.727, 0.832]), Step 8. The ranking order of the feasible alternative
(0.093, 0.957, 0.907) (0.057, 0.910, 0.785) according to the relative closeness co-efficient of the
B3 ([0.076, 0.13],[0.87, 0.941],[0.87, 0.907]), ([0.057, 0.090],[0.928, 0.972],[0.653, 0.757]),
neutrosophic cubic ideal solution is presented as follows:
( 0.093, 0.907, 0.828) (0.126, 0.910, 0.727) B2 > B3 > B1
Therefore, B2 i.e. a Basic English medium school is the
best option for the legal guardian.
C3 C4
B1
5 Conclusions
([0.132, 0.17],[0.720, 0.830],[0.625, 0.782]), ([0.124, 0.211],[0.552, 0.699],[0.552, 0.733]),
(0.084, 0.625, 0.625) (0.164, 0.660, 0.789) In the paper, we have presented a new extended TOPSIS
B2 ([0.100, 0.132],[0.720, 0.807],[0.543, 0.720]), ([0.056, 0.088],[0.66, 0.762],[0.552, 0.699]),method for solving MADM problems with neutrosophic
(0.084, 0.625, 0.625) (0.124, 0.552, 0.552) cubic information. We have proposed several operational
B3 ([0.100, 0.218],[0.625, 0.782],[0.679, 0.754]), ([0.164, 0.267],[0.552, 0.660],[0.660, 0.699]), rules on neutrosophic cubic sets. We have defined
(0.132, 0.720, 0.720) (0.088, 0.522, 0.660) Euclidean distance between two neutrosophic cubic sets.
We have defined arithmetic average operator for
neutrosophic cubic numbers. We have employed
Step 5. From Table 3, the NCPIS z jU , j = 1, 2, 3, 4 is information entropy scheme to calculate unknown weights
obtained as follows: of the attributes. We have also defined neutrosophic cubic
z1U = < ([0.043, 0.093], [0.941, 0.972], [0.941, 0.972]), positive ideal solution and neutrosophic cubic negative
ideal solution in the decision making process. Then, the
(0.075, 0.957, 0.907) >,
U
most desirable alternative is selected based on the
z 2 = < ([0.057, 0.09], [0.972, 0.986], [0.874, 0.91]), (0.057, proposed extended TOPSIS method under neutrosophic
0.972, 0.832) >, cubic environment. Finally, we have solved a numerical
z 3 = < ([0.132, 0.218], [0.625, 0.782], [0.543, 0.72], example of MADM problem regarding school selection for
U

a legal guardian to illustarte the proposed TOPSIS method.


(0.132, 0.625, 0.625)>,
We hope that the proposed TOPSIS method will be
z 4U = < [0.164, 0.267], [0.552, 0.66], [0.552, 0.699], (0.66, effective in dealing with different MADM problems such
0.552, 0.552)>; as medical diagnosis, pattern recognition, weaver selection,
The NCNIS z j , j = 1, 2, 3, 4 is determined from the supplier selection, etc in neutrosophic cubic set
L

environment.
weighted decision matrix (see Table 3) as follows:
z1L = < [0.076, 0.13], [0.87, 0.941], [0.72, 0.779], (0.093, References
0.87, 0.72)>, [1] F. Smarandache. A unifying field of logics. Neu-
z 2L = < [0.09, 0.168], [0.874, 0.972], [0.653, 0.757], (0.126, trosophy: Neutrosophic probability, set and logic,
0.91, 0.727)>, American Research Press, Rehoboth, 1998.
L [2] H. Wang, F. Smarandache, Y. Zhang, and R. Sunder-
z 3 = < [0.1, 0.132], [0.72, 0.83], [0.679, 0.782], (0.084, raman. Single valued neutrosophic sets. Multi-space
0.782, 0.83)>, and Multi-structure, 4 (2010), 410-413.
L
z 4 = < [0.056, 0.088], [0.66, 0.762], [0.66, 0.733], (0.088, [3] H. Wang, F. Smarandache, Y. Q. Zhang, R. Sunder-
raman. Interval neutrosophic sets and logic, Hexis, Ari-
0.66, 0.789)>. zona, 2005.
[4] P. Chi, P. Liu. An extended TOPSIS method for the
Step 6. The Euclidean distance measure of each alternative multiple attribute decision making problems based on
from NCPIS is obtained as follows: interval neutrosophic set. Neutrosophic Sets and Sys-
DE = 0.1232, D E = 0.1110, DE3 = 0.1200.
1 2
tems, 1 (2013), 63-70.
[5] C. L. Hwang, K. Yoon. Multiple attribute decision
Similarly, the Euclidean distance measure of each making: methods and applications, Springer, New York,
alternative from NCNIS is computed as follows: 1981.
DE = 0.0705, D E = 0.0954, DE3 = 0.0736.
1 2
[6] R. ahin, M. Yiider. A multi-criteria neutrosophic
group decision making method based TOPSIS for sup-
plier selection. Applied Mathematics and Information

Surapati Pramanik, Partha Pratim Dey, Bibhas C. Giri, Florentin Smarandache. An extended TOPSIS for multi-attribute
decision making problems with neutrosophic cu
cubic information
28 Neutrosophic Sets and Systems, Vol. 17, 2017

Sciences, submitted. [14] I. B. Tursen, Interval valued fuzzy sets based on normal
[7] Z. Zhang and C. Wu. A novel method for single-valued norms. Fuzzy Sets and Systems, 20 (2) (1986), 191-210.
neutrosophic multi-criteria decision making with in- [15] Y. B. Jun, C. S. Kim, and K. O. Yang. Cubic sets. An-
complete weight information. Neutrosophic Sets and nals of Fuzzy Mathematics and Informatics, 4(3) (2012),
Systems, 4 (2015), 35-49. 83-98.
[8] S. Broumi, J. Ye, and F. Smarandache. An extended [16] S. Pramanik, and K. Mondal. Interval neutrosophic
TOPSIS method for multiple attribute decision making multi-attribute decision-making based on grey relation-
based on interval neutrosophic uncertain linguistic vari- al analysis. Neutrosophic Sets and Systems, 9 (2015),
ables. Neutrosophic Sets and Systems, 8(2015), 23-32. 13-22.
[9] P. Biswas, S. Pramanik, and B.C. Giri. TOPSIS method [17] P. Majumder, and S.K. Samanta. On similarity and en-
for multi-attribute group decision making under single tropy of neutrosophic sets. Journal of Intelligent and
valued neutrosophic environment. Neural Computing Fuzzy Systems, 26 (2014), 12451252.
and Applications, 27 (30 (2016), 727-737. [18] C.L. Hwang, K. Yoon. Multiple attribute decision mak-
[10] M. Ali, I. Deli, and F. Smarandache. The theory of neu- ing: methods and applications: a state-of-the-art survey,
trosophic cubic sets and their applications in pattern Springer, London, 1981.
recognition. Journal of Intelligent and Fuzzy Systems, [19] J.Q. Wang, and Z.H Zhang. Multi-criteria decision
30 (4) (2016), 1957-1963. making method with incomplete certain information
[11] Y.B. Jun, F. Smarandache, C.S. Kim. Neutrosophic based on intuitionistic fuzzy number. Control and
cubic sets. New Mathematics and Natural Decision, 24 (2009), 226-230.
Computation, 13 (1) (2017), 41-54. [20] K. Mondal, and S. Pramanik. Neutrosophic decision
[12] S. Pramanik, S. Dalapati, S. Alam, T. K. Roy, F making model of school choice. Neutrosophic Sets and
Smarandache. Neutrosophic cubic MCGDM method Systems, 7(2015), 62-68.
based on similarity measure. Neutrosophic Sets and
Systems 16 (2017), 44-56.

[13] L.A. Zadeh, Fuzzy sets. Information and Control, 8


(1965), 338-353. Received: July 13, 2017. August 1, 2017.

________________________________________________________________________________________________
Surapati Pramanik, Partha Pratim Dey, Bibhas C. Giri, Florentin Smarandache. An extended TOPSIS for multi-attribute
decision making problems with neutrosophic cubic information