Professional Documents
Culture Documents
DOI:10.3233/JIFS-161548
IOS Press
Abstract. This paper presents three novel single-valued neutrosophic soft set (SVNSS) methods. First, we initiate a new
axiomatic definition of single-valued neutrosophic similarity measure, which is expressed by single-valued neutrosophic
number (SVNN) that will reduce the information loss and remain more original information. Then, the objective weights of
various parameters are determined via grey system theory. Moreover, we develop the combined weights, which can show
both the subjective information and the objective information. Later, we propose three algorithms to solve single-valued
neutrosophic soft decision making problem by Evaluation based on Distance from Average Solution (EDAS), similarity
measure and level soft set. Finally, the effectiveness and feasibility of approaches are demonstrated by a numerical example.
Keywords: Single-valued neutrosophic soft set, similarity measure, SVNN, Combined weighed , EDAS , level soft set
1064-1246/17/$35.00 © 2017 – IOS Press and the authors. All rights reserved
956 X. Peng and C. Liu / Algorithms for neutrosophic soft decision making based on EDAS
and tautological set [17]. From scientific or engi- ronment. Mukherjee and Sarkar [43] defined some
neering point of view, the neutrosophic set and similarity measures of single-valued neutrosophic
set-theoretic operators need to be specified. Oth- soft sets, and applied them to real life problems. Deli
erwise, it will be difficult to apply in the real and Broumi [44] defined some new operators and
applications. Hence, Smarandache [17] firstly intro- soft matrix based on single-valued neutrosophic soft
duced the a single valued neutrosophic set (SVNS). sets. Alkhazaleh [45] introduced time-neutrosophic
Wang et al. [18] provided the set-theoretic operators soft set and studied some of its properties in detail.
and various properties of SVNSs. Ye [19, 20] pro- Al-Quran and Hassan [46] presented the neutosophic
posed a multi-attribute decision making (MADM) vague soft expert set theory and discussed their prop-
method using the correlation coefficient under single- erties in detail.
valued neutrosophic environment. Ye [21, 22] further Evaluation based on Distance from Average Solu-
developed clustering method and decision making tion (EDAS), originally proposed by Ghorabaee et al.
methods by similarity measures of SVNS. Mean- [47], is a new MADM method for inventory ABC
while, Peng and Dai [23] presented new similarity classification. It is very useful when we have some
measure of SVNS and applied them to decision mak- conflicting parameters. In the compromise MADM
ing. Biswas et al. [24] extended the Technique for methods such as TOPSIS and VIKOR [48], the best
Order Preference by Similarity to an Ideal Solution alternative is got by computing the distance from
(TOPSIS) method for multi-attribute single-valued ideal and nadir solutions. The desirable alternative
neutrosophic decision-making problem. Sahin and has lower distance from ideal solution and higher dis-
Kucuk [25] defined a subsethood measure for SVNS, tance from nadir solution in these MADM methods.
and applied to MADM. Yang et al. [26] introduced Ghorabaee et al. [49] extended the EDAS method to
single valued neutrosophic relations and discussed supplier selection. As far as we know, however, the
their properties. Huang [27] developed a new distance study of the MADM problem based on EDAS method
measure of SVNSs, and applied them to clustering have not been reported in the existing academic lit-
analysis and MADM. Liu [28] proposed aggrega- erature. Hence, it is an interesting research topic to
tion operators based on Archimedean t-conorm and apply the EDAS in MADM to rank and determine the
t-norm for SVNS and also gave an application in best alternative under single-valued neutrosophic soft
MADM. Wu et al. [29] proposed a cross-entropy environment. Through a comparison analysis of the
and prioritized aggregation operator with simpli- given methods, their objective evaluation is carried
fied neutrosophic sets. Single valued neutrosophic out, and the method which maintains consistency of
graphs [30–32] and bipolar single valued neutro- its results is chosen.
sophic graphs [33, 34] are proposed by Broumi et For computing the similarity measure of two
al. Peng et al. [35, 36] proposed ELECTRE approach SVNSs, we propose a new axiomatic definition of
and qualitative flexible approach based on likelihood similarity measure, which takes in the form of SVNN.
for multi-valued neutrosophic MADM problem, Comparing with the existing literature [21, 22, 42,
respectively. Ji et al. [37] introduced a projection- 43], our similarity measure can remain more original
based TODIM method for multi-valued neutrosophic decision information.
MADM problem. Tian et al. [38] presented an By means of level soft sets, Feng et al. [7] pre-
improved MULTIMOORA approach for simpli- sented an adjustable approach to fuzzy soft sets based
fied neutrosophic linguistic set. Later, Tian et al. decision making. By considering different types of
[39] incorporated power aggregation operators and thresholds, it can derive different level soft sets from
a TOPSIS-based QUALIFLEX method in order to the original fuzzy soft set. In general, the final opti-
solve green product design selection problems using mal decisions based on different level soft sets could
simplified neutrosophic linguistic information. Tian be different. Thus the newly proposed approach is in
et al. [40] proposed a MADM based on generalized fact an adjustable method which captures an impor-
prioritized aggregation operators under simplified tant feature for decision making in an imprecise
neutrosophic uncertain linguistic environment. environment: some of these problems are essentially
Combining the advantages of the instance of neu- humanistic and thus subjective in nature. As far as we
trosophic set [17] with soft set [1], Maji [41] proposed know, however, the study of the single-valued neu-
the single-valued neutrosophic soft set. Şahin and trosophic soft MADM problem based on level soft
Küçük [42] presented the definition of similarity set have not been reported in the existing academic
and entropy in single-valued neutrosophic soft envi- literature.
X. Peng and C. Liu / Algorithms for neutrosophic soft decision making based on EDAS 957
Considering that different attribute weights will falsity-membership function FB (x). The functions
influence the ranking results of alternatives, we TB (x), IB (x), and FB (x) are real standard or non-
develop a new method to determine the attribute standard subsets of ]0− , 1+ [ . That is TB (x) : X →
weights by combining the subjective elements with ]0− , 1+ [ , IB (x) : X →]0− , 1+ [ , and FB (x) : X →
the objective ones. This model is different from the ]0− , 1+ [.
existing methods, which can be divided into two There is restriction on the sum of TB (x), IB (x),
tactics: one is the subjective weighting evaluation and FB (x), so 0− ≤ sup TB (x) + sup IB (x) +
methods and the other is the objective weighting sup FB (x) ≤ 3+ .
determine methods, which can be computed by grey As mentioned above, it is hard to apply the neu-
system theory [50]. The subjective weighting meth- trosophic set to solve some real problems. Hence,
ods focus on the preference information of the Smarandache [17] presented SVNS, which is a sub-
decision maker [19–21, 27–29], while they ignore class of the neutrosophic set and mentioned the
the objective information. The objective weighting definition as follows:
determine methods do not take the preference of the
decision maker into account, that is to say, these meth- Definition 2. [17] Let X be a universe of dis-
ods fail to take the risk attitude of the decision maker course, with a class of elements in X denoted by
into account [22, 24]. The feature of our weighting x. A single-valued neutrosophic set N in X is sum-
model can show both the subjective information and marized by a truth-membership function TN (x), an
the objective information. Hence, combining objec- indeterminacy-membership function IN (x), and a
tive weights with subjective weights, a combined falsity-membership function FN (x). Then a SVNS N
model to obtain attribute weights is proposed. can be denoted as follows:
The remainder of this paper is organized as fol-
lows: In Section 2, we review some fundamental N = {< x, TN (x), IN (x), FN (x) >| x ∈ X}, (1)
conceptions of neutrosophic sets, single-valued neu-
where TN (x), IN (x), FN (x) ∈ [0, 1] for ∀x ∈ X.
trosophic sets, soft set and single-valued neutrosophic
Meanwhile, the sum of TN (x), IN (x), and FN (x) ful-
soft sets. In Section 3, a new axiomatic definition of
fills the condition 0 ≤ TN (x) + IN (x) + FN (x) ≤ 3.
single-valued neutrosophic similarity measure is pre-
For a SVNS N in X, the triplet (TN (x), IN (x), FN (x))
sented. In Section 4, three single-valued neutrosophic
is called single-valued neutrosophic number
soft decision making approaches based on EDAS,
(SVNN). For convenience, we can simply use
similarity measure and level soft set are shown. In
x = (Tx , Ix , Fx ) to represent a SVNN as an element
Section 5, a numerical example is given to illustrate
in the SVNS N.
the proposed methods. The paper is concluded in
Generally speaking, two special values are taken
Section 6.
into consideration, i.e., SVNN 0 and 1. If we think
about 0 as the worst value and 1 as the best value, we
can set 0 as (0,1,1) and 1 as (1,0,0).
2. Preliminaries
Definition 3. [17, 18] Let x = (Tx , Ix , Fx ) and y =
2.1. Neutrosophic set (Ty , Iy , Fy ) be two SVNNs, then operations can be
defined as follows:
Neutrosophic set is a portion of neutrosophy, which
researches the origin, and domain of neutralities, (1) xc= (Fx , 1 − Ix , Tx );
as well as their interactions with diverse ideational (2) x y = (max{Tx , Ty }, min{Ix , Iy }, min
scope [17], and is a convincing general formal frame- {Fx , Fy });
work, which extends the presented sets [2, 4] from (3) x y = (min{Tx , Ty }, max{Ix , Iy }, max
philosophical point. Smarandache [17] introduced {Fx , Fy });
the definition of neutrosophic set as follows: (4) x ⊕ y = (Tx + Ty − Tx ∗ Ty , Ix ∗ Iy , Fx ∗
Fy );
Definition 1. [17] Let X be a universe of dis- (5) x ⊗ y = (Tx ∗ Ty , Ix + Iy − Ix ∗ Iy , Fx +
course, with a class of elements in X denoted Fy − Fx ∗ Fy );
by x. A neutrosophic set B in X is summa- (6) λx = (1 − (1 − Tx )λ , (Ix )λ , (Fx )λ ), λ > 0;
rized by a truth-membership function TB (x), an (7) xλ = ((Tx )λ , 1 − (1 − Ix )λ , 1 − (1 −
indeterminacy-membership function IB (x), and a Fx )λ ), λ > 0.
958 X. Peng and C. Liu / Algorithms for neutrosophic soft decision making based on EDAS
n
It means that L(Ai , Ak ) = M(Ai , Ak ) = R(Ai , Ak )
wj min{1−Iij ,1−Ikj }
= 1, W(Ai , Ak ) = 0.
M(Ai , Ak ) =
j=1
n , Furthermore,
wj max{1−Iij ,1−Ikj }
n
j=1 wj min{Tij ,Tkj }
n
L(Ai , Ak ) =
j=1
= 1,
wj min{1−Iij ,1−Ikj } n
wj max{Tij ,Tkj }
W(Ai , Ak ) = 1 −
j=1
n , j=1
wj max{1−Iij ,1−Ikj } n
j=1 wj min{1−Iij ,1−Ikj }
n
M(Ai , Ak ) =
j=1
= 1,
wj min{1−Fij ,1−Fkj } n
wj max{1−Iij ,1−Ikj }
R(Ai , Ak ) =
j=1
n and wj is the j=1
wj max{1−Fij ,1−Fkj }
n
j=1 wj min{1−Iij ,1−Ikj }
weight of jth SVNN. W(Ai , Ak ) = 1 −
j=1
= 0,
n
wj max{1−Iij ,1−Ikj }
Proof. In order for S (Ai , Ak ) to be qualified as a j=1
sensible similarity measure for SVNSs, it must satisfy
n
wj min{1−Fij ,1−Fkj }
the (1)-(4) of axiomatic requirements.
n R(Ai , Ak ) =
j=1
n = 1.
wj min{Tij ,Tkj }
wj max{1−Fij ,1−Fkj }
(1) Since 0 ≤ L(Ai , Ak ) = ≤ 1,
j=1
n j=1
wj max{Tij ,Tkj } Based on the randomicity of wj , we can have Tij =
j=1 Tkj , Iij = Ikj , Fij = Fkj , i.e., Ai = Ak .
n
② Sufficiency:
wj min{1−Iij ,1−Ikj }
0 ≤ M(Ai , Ak ) =
j=1
≤ 1, Since Ai = Ak , we have Tij = Tkj , Iij = Ikj , Fij =
n
wj max{1−Iij ,1−Ikj } Fkj .
j=1 Furthermore,
n
wj min{1−Iij ,1−Ikj }
0 ≤ W(Ai , Ak ) = 1 −
j=1
≤ 1, wj min{Tij , Tkj } = wj max{Tij , Tkj },
n
wj max{1−Iij ,1−Ikj } j=1 j=1
j=1
n
n
wj min{1 − Iij , 1 − Ikj }
wj min{1−Fij ,1−Fkj }
j=1
0 ≤ R(Ai , Ak ) = ≤ 1,
j=1
n
wj max{1−Fij ,1−Fkj }
n
j=1 = wj max{1 − Iij , 1 − Ikj },
therefore, j=1
0 ≤ min{L(Ai , Ak ), M(Ai , Ak ), R(Ai , Ak )} ≤ 1,
0 ≤ min{L(Ai , Ak ), W(Ai , Ak ), R(Ai , Ak )} ≤ 1,
n
wj min{1 − Fij , 1 − Fkj }
0 ≤ 1 − max{L(Ai , Ak ), M(Ai , Ak ), R(Ai , Ak )}
j=1
≤ 1.
Furthermore,
n
= wj max{1 − Fij , 1 − Fkj }.
j=1
0 ≤ min{L(Ai , Ak ), M(Ai , Ak ), R(Ai , Ak )}
+ min{L(Ai , Ak ), W(Ai , Ak ), R(Ai , Ak )} + 1 Consequently, S (Ai , Ak ) = (1, 0, 0).
(3) It is obvious.
− max{L(Ai , Ak ), M(Ai , Ak ), R(Ai , Ak )} ≤ 3. (4) If A1 ⊆ A2 ⊆ A3 , then ∀j, T1j ≤ T2j ≤
T3j , I1j ≥ I2j ≥ I3j and F1j ≥ F2j ≥ F3j .
Consequently, S (Ai , Ak ) is a SVNN. Hence,
(2) ① Necessity:
n
Since S (Ai , Ak ) = (1, 0, 0), we have wj min{T1j , T3j }
min{L(Ai , Ak ), M(Ai , Ak ), R(Ai , Ak )} = 1, L(A1 , A3 ) =
j=1
min{L(Ai , Ak ), W(Ai , Ak ), R(Ai , Ak )} = 0, n
wj max{T1j , T3j }
max{L(Ai , Ak ), M(Ai , Ak ), R(Ai , Ak )} = 1. j=1
960 X. Peng and C. Liu / Algorithms for neutrosophic soft decision making based on EDAS
n
n
n
n
wj T1j wj T1j wj (1 − F1j ) wj (1 − F1j )
j=1 j=1 j=1 j=1
= ≤ = ≤
n n n n
wj T3j wj T2j wj (1 − F3j ) wj (1 − F2j )
j=1 j=1 j=1 j=1
n
n
wj min{T1j , T2j } wj min{1 − F1j , 1 − F2j }
j=1 j=1
= = L(A1 , A2 ), =
n n
wj max{T1j , T2j } wj max{1 − F1j , 1 − F2j }
j=1 j=1
n
= R(A1 , A2 ).
wj min{1 − I1j , 1 − I3j }
j=1
M(A1 , A3 ) = Furthermore,
n
wj max{1 − I1j , 1 − I3j }
j=1 min{L(A1 , A2 ), M(A1 , A2 ), R(A1 , A2 )}
n
n ≥ min{L(A1 , A3 ), M(A1 , A3 ), R(A1 , A3 )},
wj (1 − I1j ) wj (1 − I1j )
j=1 j=1 min{L(A1 , A2 ), W(A1 , A2 ), R(A1 , A2 )}
= ≤
n n
wj (1 − I3j ) wj (1 − I2j ) ≥ min{L(A1 , A3 ), W(A1 , A3 ), R(A1 , A3 )},
j=1 j=1
n 1 − max{L(A1 , A2 ), M(A1 , A2 ), R(A1 , A2 )}
wj min{1 − I1j , 1 − I2j }
j=1 ≤ 1 − max{L(A1 , A2 ), M(A1 , A2 ), R(A1 , A2 )}.
=
n
Consequently, S (A1 , A2 ) ⊇ S (A1 , A3 ).
wj max{1 − I1j , 1 − I2j }
j=1 Similarly, S (A2 , A3 ) ⊇ S (A1 , A3 ). This com-
= M(A1 , A2 ), pletes the proof.
n
Example 2. [21] Assume that we have the follow-
wj min{1 − I1j , 1 − I3j }
j=1 ing three SVNSs in a universe of discourse X =
W(A1 , A3 ) = 1 − {x1 , x2 }, A ⊆ B ⊆ C, w = {0.5, 0.5}:
n
wj max{1 − I1j , 1 − I3j }
j=1 A = {< x1 , 0.1, 0.5, 0.6 >, < x2 , 0.2, 0.5, 0.7 >},
n
wj (1 − I1j ) B = {< x1 , 0.3, 0.4, 0.5 >, < x2 , 0.5, 0.3, 0.4 >},
j=1
= 1− C = {< x1 , 0.6, 0.1, 0.2 >, < x2 , 0.8, 0.1, 0.3 >}.
n
wj (1 − I3j )
j=1 By applying the proposed similarity measure,
n
S (A, B) = (0.3750, 0.2308, 0.2308),
wj (1 − I1j )
j=1
≤ 1− S (B, C) = (0.5714, 0.2778, 0.2778),
n
wj (1 − I2j ) S (A, C) = (0.2143, 0.2143, 0.4444).
j=1
n Thus, S (A, C) < S (A, B) and S (A, C) <
wj min{1 − I1j , 1 − I2j } S (B, C).
j=1
= 1−
n
wj max{1 − I1j , 1 − I2j }
j=1 4. Three algorithms for single-valued
= W(A1 , A2 ), neutrosophic soft decision making
n
wj min{1 − F1j , 1 − F3j } 4.1. Problem description
j=1
R(A1 , A3 ) =
n
Let U = {x1 , x2 , · · · , xm } be a finite set of m
wj max{1 − F1j , 1 − F3j }
j=1 alternatives, E = {e1 , e2 , · · · , en } be a set of n
X. Peng and C. Liu / Algorithms for neutrosophic soft decision making based on EDAS 961
⎧
operations of the SVNN are utilized for extending the ⎪ (ej )(xi ))−s(AVj )}
max{0,s(F
⎨ s(AVj ) , ej ∈ B,
EDAS method.
Pij = (ej )(xi ))} (10)
⎪
⎩ max{0,s(AVs(AV
j )−s(F
, ej ∈ C,
Algorithm 1: EDAS j)
⎧
Step 1. Identify the alternatives and parameters, and ⎪ (ej )(xi ))}
max{0,s(AVj )−s(F
, E) ⎨ s(AVj ) , ej ∈ B,
obtain the single-valued neutrosophic soft set (F
Nij = (11)
which is shown in Table 2. ⎩ max{0,s(F(e
⎪ j )(xi ))−s(AVj )}
s(AVj ) , ej ∈ C,
Step 2. Normalize the single-valued neutrosophic
, E) into (F
soft set (F
, E) by Equation (7). where s(AVj ) and s(F (ej )(xi )) are score function of
AVj and F (ej )(xi ), respectively.
(ej )(xi )
F
⎧ Step 6. Determine the weighted sum of PDA and
⎪
⎨ T (e )(x ), I (e )(x ), F (e )(x ) , ej ∈ B, NDA for all alternatives, shown as follows:
F j i
F j i F j i
= (7)
⎪
⎩ F (e )(xi ), 1 − I (e )(xi ), T (e )(xi ) , ej ∈ C,
n
F j F j F j SPi = wj Pij , (12)
j=1
n
eter set. SNi = wj Nij . (13)
j=1
Step 3. Compute the combined weights by Equa-
tion (6). Step 7. Normalize the values of SPi and SNi for all
alternatives, shown as follows:
Step 4. Determine the average solution according to
all parameters, shown as follows: SPi
NSPi = (14)
max{SPi }
i
AV = (AVj )1×n , (8)
SNi
NSNi = 1 − (15)
where max{SNi }
i
1
m
AVj = F (ej )(xi ) Step 8. Calculate the appraisal score ASi (i =
m
i=1 1, 2, · · · , m) for all alternatives, shown as follows:
m 1
1 ASi = (NSPi + NSNi ),
= 1− (1 − TF (ej )(xi )), 2
(16)
m
i=1
where 0 ≤ ASi ≤ 1.
m
m
IF (ej )(xi ), FF (ej )(xi ) Step 9. Rank the alternatives by means of the decreas-
i=1 i=1
ing values of ASi . The alternative with the highest ASi
m 1 is the best choice among the candidate alternatives.
m
= 1− (1 − TF (ej )(xi )) ,
i=1
4.4. The method of similarity measure
m 1
m
m 1
IF (ej )(xi ) , ( FF (ej )(xi )) m . In this section, we introduce a method for the
i=1 i=1 decision making problem by the proposed similarity
(9) measure between SVNSs. The concept of ideal point
has been applied to help determine the best alternative
Step 5. Calculate the positive distance from average in the decision process. Although the ideal alternative
(PDA) with PDA = (Pij )m×n and the negative dis- does not exist in practical problems, it does offer a
tance from average (NDA) with NDA = (Nij )m×n useful theoretical construct against which to appraise
matrixes according to the type of parameters, shown alternatives. Therefore, we define the ideal alternative
as follows: x∗ as the SVNN xj∗ = (T ∗ , I ∗ , F ∗ ) = (1, 0, 0) for ∀j.
X. Peng and C. Liu / Algorithms for neutrosophic soft decision making based on EDAS 963
Hence, by applying Equation (3), the proposed Step 4. Calculate the similarity measure S(xi , x∗ )
similarity measure S between an alternative xi and (i = 1, 2, · · · , m) by Equation (17).
the ideal alternative x∗ represented by the SVNSs is
defined by Step 5. Compute the each alternative of score func-
tion s(S(xi , x∗ )) by Equation (2).
S (xi , x∗ ) Step 6. Rank the alternatives by s(S(xi , x∗ )) (i =
1, 2, · · · , m). The most desired alternative is the one
= min{L(xi , x∗ ), M(xi , x∗ ), R(xi , x∗ )}, with the biggest value of xi .
min{L(xi , x∗ ), W(xi , x∗ ), R(xi , x∗ )}, 4.5. The method of level soft set
1 − max{L(xi , x∗ ), M(xi , x∗ ), R(xi , x∗ )} , (17) We present an adjustable approach to single-valued
neutrosophic soft set based decision making prob-
where lems. This proposal is based on the following novel
concept called level soft sets.
n
j min{TF (ej )(xi ), 1} Definition 10. Let = (F , E) be a SVNSS over
L(xi , x∗ ) =
j=1 U. Let λ : E → [0, 1] be a function, i.e. ∀ej ∈ E,
n
λ(ej ) = (r(ej ), k(ej ), t(ej )), and r(ej ), k(ej ), t(ej ) ∈
j max{TF (ej )(xi ), 1}
j=1 [0, 1]. The level soft set of with respect to λ is
a crisp soft set L(; λ) =< F λ , E > defined by
n
= j TF (ej )(xi ), Fλ (ej ) = L(F (ej ); λ(ej )) = {xi ∈ U|TF (ej )(xi ) ≥
j=1
r(ej ), IF (ej )(xi ) ≥ k(ej ), FF (ej )(xi ) ≤ t(ej )}, for all
ej ∈ E(j = 1, 2, · · · , n), xi ∈ U(i = 1, 2, · · · , m).
n
j min{1 − IF (ej )(xi ), 1} Remark 1. In the above definition, the function
∗ j=1
M(xi , x ) = λ : E → [0, 1]3 is not restricted to be SVNS, it is
n
j max{1 − IF (ej )(xi ), 1} only a function, r(ej ) ∈ [0, 1] can be viewed as a
j=1 given least threshold (the parameter ej on the degree
n
= j IF (ej )(xi ), function defined in Equation (2).
j=1
For convenience, we choose mid-level threshold
function mid E → [0, 1]3 , i.e.,
n
mid (ej ) = (rmid (ej ), kmid (ej ), tmid (ej )) for all
j min{1 − FF (ej )(xi ), 1}
j=1 ej ∈ E, where
R(xi , x∗ ) =
n
1
m
j max{1 − FF (ej )(xi ), 1}
j=1 rmid (ej ) = TF (ej )(xi ),
m
i=1
n
= 1− j FF (ej )(xi ). 1
n
ci = i L(ej )(xi ). (18) (6) as follows:
j=1
1 = 0.4078, 2 = 0.1017, 3 = 0.4905.
Step 6. The optimal decision is to select xk if ck =
m Step 4. Determine the average solution according to
max ci .
i=1 all parameters by Equation (9), shown as follows:
Table 6 Acknowledgments
A comparison study with some existing methods
Method The final ranking The optimal The authors are very appreciative to the review-
alternative
ers for their precious comments which enormously
Deli and Broumi [44] x3 x4 x1 x2 x3 ameliorated the quality of this paper. Our work is
Maji [53] x3 x4 x1 x2 x3
Karaaslan [54] x3 x4 x1 x2 x3 sponsored by the National Natural Science Founda-
Deli and Broumi [55] x3 x4 x1 x2 x3 tion of China (Nos. 61163036, 61462019).
Algorithm 1 x3 x2 x4 x1 x3
Algorithm 2 x3 x4 x1 x2 x3
Algorithm 3 x3 x4 x2 x1 x3 References
[19] J. Ye, Multicriteria decision-making method using the Neural Computing and Applications. doi: 10.1007/s00521-
correlation coefficient under single-valued neutrosophic 016-2411-8
environment, International Journal of General Systems 42 [37] P. Ji, H.Y. Zhang and J.Q. Wang, A projection-based
(2013), 386–394. TODIM method under multi-valued neutrosophic environ-
[20] J. Ye, Improved correlation coefficients of single valued neu- ments and its application in personnel selection, Neural
trosophic sets and interval neutrosophic sets for multiple Computing and Applications. doi: 10.1007/s00521-016-
attribute decision making, Journal of Intelligent and Fuzzy 2436-z
Systems 27 (2014), 2453–2462. [38] Z.P. Tian, J. Wang, J.Q. Wang and H.Y. Zhang, An
[21] J. Ye, Clustering methods using distance-based similarity improved MULTIMOORA approach for multi- criteria
measures of single-valued neutrosophic sets, Journal of decision-making based on interdependent inputs of simpli-
Intelligent Systems 23 (2014), 379–389. fied neutrosophic linguistic information, Neural Computing
[22] J. Ye, Multiple attribute group decision-making method with and Applications. doi: 10.1007/s00521-016-2378-5
completely unknown weights based on similarity measures [39] Z.P. Tian, J. Wang, J.Q. Wang and H.Y. Zhang,
under single valued neutrosophic environment, Journal of Simplified neutrosophic linguistic multi-criteria group
Intelligent and Fuzzy Systems 27 (2014), 2927–2935. decision-making approach to green product development,
[23] X.D. Peng and J.G. Dai, Approaches to single-valued neu- Group Decision and Negotiation. doi: 10.1007/s10726-016-
trosophic MADM based on MABAC, TOPSIS and new 9479-5
similarity measure with score function, Neural Computing [40] Z.P. Tian, J. Wang, H.Y. Zhang and J.Q. Wang, Multi-criteria
and Applications. doi:10.1007/s00521-016-2607-y decision-making based on generalized prioritized aggre-
[24] P. Biswas, S. Pramanik and B.C. Giri, TOPSIS method for gation operators under simplified neutrosophic uncertain
multi-attribute group decision-making under single-valued linguistic environment, International Journal of Machine
neutrosophic environment, Neural Computing and Applica- Learning and Cybernetics. doi: 10.1007/s13042-016-
tions 27 (2016), 727–737. 0552-9
[25] R. Sahin and A. Kucuk, Subsethood measure for single [41] P.K. Maji, Neutrosophic soft set, Annals of Fuzzy Mathe-
valued neutrosophic sets, Journal of Intelligent and Fuzzy matics and Informatics 5 (2013), 157–168.
Systems 29 (2015), 525–530. [42] R. Şahin and A. Küçük, On similarity and entropy of neu-
[26] H.L. Yang, Z.L. Guo, Y.H. She and X.W. Liao, On single val- trosophic soft sets, Journal of Intelligent and Fuzzy Systems
ued neutrosophic relations, Journal of Intelligent and Fuzzy 27 (2014), 2417–2430.
Systems 30 (2016), 1045–1056. [43] A. Mukherjee and S. Sarkar, Several similarity measures of
[27] H.L. Huang, New distance measure of single-valued neu- neutrosophic soft sets and its application in real life prob-
trosophic sets and its application, International Journal of lems, Annals of Pure and Applied Mathematics 7 (2014),
Intelligent Systems 31 (2016), 1021–1032. 1–6.
[28] P.D. Liu, The aggregation operators based on archimedean [44] I. Deli and S. Broumi, Neutrosophic soft matrices and NSM
t-Conorm and t-Norm for single-valued neutrosophic num- decision making, Journal of Intelligent and Fuzzy Systems
bers and their application to decision making, International 28 (2015), 2233–2241.
Journal of Fuzzy Systems. doi:10.1007/s40815-016-0195-8 [45] S. Alkhazaleh, Time-neutrosophic soft set and its applica-
[29] X.H.Wu, J. Wang, J.J. Peng and X.H. Chen, Cross- tions, Journal of Intelligent and Fuzzy Systems 30 (2016),
entropy and prioritized aggregation operator with simplified 1087–1098.
neutrosophic sets and their application in multi-criteria [46] A. Al-Quran and N. Hassan, Neutrosophic vague soft expert
decisionmaking problems, International Journal of Fuzzy set theory, Journal of Intelligent and Fuzzy Systems 30
Systems. doi: 10.1007/s40815-016-0180-2 (2016), 3691–3702.
[30] S. Broumi, M. Talea, A. Bakali and F. Smarandache, Sin- [47] M.K. Ghorabaee, E.K. Zavadskas, L. Olfat and Z. Turskis,
gle valued neutrosophic graphs, Journal of New Theory 10 Multi-criteria inventory classification using a new method of
(2016), 86–101. evaluation based on distance from average solution (EDAS),
[31] S. Broumi, A. Bakali, M. Talea and F. Smarandache, Isolated Informatica 26 (2015), 435–451.
single valued neutrosophic graphs, Neutrosophic Sets and [48] S. Opricovic and G.H. Tzeng, Compromise solution by
Systems 11 (2016), 74–78. MCDMmethods: A comparative analysis of VIKOR and
[32] S. Broumi, M. Talea, F. Smarandache and A. Bakali, Single TOPSIS, European Journal of Operational Research 156
valued neutrosophic graphs: Degree, order and size, IEEE (2004), 445–455.
World Congress on Computational Intelligence (2016), [49] M.K. Ghorabaee, E.K. Zavadskas, M. Amiri and Z. Turskis,
2444–2451. Extended EDAS method for fuzzy multi-criteria decision-
[33] S. Broumi, M. Talea, A. Bakali and F. Smarandache, On making: An application to supplier selection, International
bipolar single valued neutrosophic graphs, Journal of New Journal of Computers Communications and Control 11
Theory 11 (2016), 84–102. (2016), 358–371.
[34] S. Broumi, F. Smarandache, M. Talea and A. Bakali, An [50] S.F. Liu, Y.G. Dang and Z.G. Fang, Grey systems theory and
introduction to bipolar single valued neutrosophic graph its applications, Science Press, Beijing, 2000.
theory, Applied Mechanics and Materials 841 (2016), [51] J.J. Peng, J.Q. Wang, J. Wang, H.Y. Zhang and X.H.
184–191. Chen, Simplified neutrosophic sets and their applica-
[35] J.J. Peng, J.Q. Wang and W.E. Yang, A multi-valued neu- tions in multicriteria group decision-making problems,
trosophic qualitative flexible approach based on likelihood International Journal of Systems Science 47 (2016),
for multi-criteria decision-making problems, International 2342–2358.
Journal of Systems Science. doi: 10.1080/00207721. [52] F. Smarandache, Neutrosophic Overset, Neutrosophic
2016.1218975 Underset, and Neutrosophic Offset. Similarly for Neutro-
[36] J.J. Peng, J.Q. Wang and X. Wu, An extension of the ELEC- sophic Over-/Under-/Off- Logic, Probability, and Statistics.
TRE approach with multi-valued neutrosophic information, Pons Editions, Brussels, Belgium, 2016.
968 X. Peng and C. Liu / Algorithms for neutrosophic soft decision making based on EDAS
[53] P.K. Maji, A neutrosophic soft set approach to a decision [55] I. Deli and S. Broumi, Neutrosophic soft relations and some
making problem, Annals of Fuzzy Mathematics and Infor- properties, Annals of Fuzzy Mathematics and Informatics 9
matics 3 (2012), 313–319. (2015), 169–182.
[54] F. Karaaslan, Neutrosophic soft sets with applications in
decision making, International Journal of Information Sci-
ence and Intelligent System 4 (2015), 1–20.