SS symmetry Article
Models for Green Supplier Selection with Some 2-Tuple Linguistic Neutrosophic Number Bonferroni Mean Operators Jie Wang 1 , Guiwu Wei 1, * 1 2
*
ID
and Yu Wei 2, *
School of Business, Sichuan Normal University, Chengdu 610101, China; JW970326@163.com School of Finance, Yunnan University of Finance and Economics, Kunming 650221, China Correspondence: weiguiwu1973@sicnu.edu.cn (G.W.); ywei@home.swjtu.edu.cn (Y.W.)
Received: 4 April 2018; Accepted: 16 April 2018; Published: 25 April 2018
Abstract: In this paper, we extend the Bonferroni mean (BM) operator, generalized Bonferroni mean (GBM) operator, dual generalized Bonferroni mean (DGBM) operator and dual generalized geometric Bonferroni mean (DGGBM) operator with 2-tuple linguistic neutrosophic numbers (2TLNNs) to propose 2-tuple linguistic neutrosophic numbers weighted Bonferroni mean (2TLNNWBM) operator, 2-tuple linguistic neutrosophic numbers weighted geometric Bonferroni mean (2TLNNWGBM) operator, generalized 2-tuple linguistic neutrosophic numbers weighted Bonferroni mean (G2TLNNWBM) operator, generalized 2-tuple linguistic neutrosophic numbers weighted geometric Bonferroni mean (G2TLNNWGBM) operator, dual generalized 2-tuple linguistic neutrosophic numbers weighted Bonferroni mean (DG2TLNNWBM) operator, and dual generalized 2-tuple linguistic neutrosophic numbers weighted geometric Bonferroni mean (DG2TLNNWGBM) operator. Then, the MADM methods are proposed with these operators. In the end, we utilize an applicable example for green supplier selection in green supply chain management to prove the proposed methods. Keywords: multiple attribute decision making (MADM); neutrosophic numbers; 2-tuple linguistic neutrosophic numbers set (2TLNNSs); Bonferroni mean (BM) operator; generalized Bonferroni mean (GBM) operator; dual generalized Bonferroni mean (DGBM) operator; dual generalized geometric Bonferroni mean (DGGBM) operator; green supplier selection; green supply chain management
1. Introduction Zadeh [1] introduced a membership function between 0 and 1 instead of traditional crisp value of 0 and 1, and defined the fuzzy set (FS). In order to overcome the insufficiency of FS, Atanassov [2] proposed the concept of an intuitionistic fuzzy set (IFS), which is characterized by its membership function and non-membership function between 0 and 1. Furthermore, Atanassov and Gargov [3] introduced the concept of an interval-valued intuitionistic fuzzy set (IVIFS), which is characterized by its interval membership function and interval non-membership function in the unit interval [0,1]. Because IFSs and IVIFSs cannot depict indeterminate and inconsistent information, Smarandache [4] introduced a neutrosophic set (NS) from a philosophical point of view to express indeterminate and inconsistent information. A NS has more potential power than other modeling mathematical tools, such as fuzzy set [1], IFS [2], and IVIFS [3]. But, it is difficult to apply NSs in solving of real life problems. Therefore, Smarandache [4] and Wang et al. [5,6] defined a single valued neutrosophic set (SVNS) and an interval neutrosophic set (INS), which are characterized by a truth-membership, an indeterminacy membership, and a falsity membership. Ye [7] introduced a simplified neutrosophic set (SNS), including the concepts of SVNS and INS, which are the extension of IFS and IVIFS. Obviously,
Symmetry 2018, 10, 131; doi:10.3390/sym10050131
www.mdpi.com/journal/symmetry
Symmetry 2018, 10, 131
2 of 36
SNS is a subclass of NS, while SVNS and INS are subclasses of SNS. Ye [8] proposed the correlation and correlation coefficient of single-valued neutrosophic sets (SVNSs) that are based on the extension of the correlation of intuitionistic fuzzy sets and demonstrates that the cosine similarity measure is a special case of the correlation coefficient in SVNS. Broumi and Smarandache [9] extended the correlation coefficient to INSs. Biswas et al. [10] developed a new approach for multi-attribute group decision-making problems by extending the technique for order preference by similarity to ideal solution to single-valued neutrosophic environment. Liu et al. [11] combined Hamacher operations and generalized aggregation operators to NSs, and proposed the generalized neutrosophic number Hamacher weighted averaging (GNNHWA) operator, generalized neutrosophic number Hamacher ordered weighted averaging (GNNHOWA) operator, and generalized neutrosophic number Hamacher hybrid averaging (GNNHHA) operator, and explored some properties of these operators and analyzed some special cases of them. Sahin and Liu [12] developed a maximizing deviation method for solving the multiple attribute decision-making problems with the single-valued neutrosophic information or interval neutrosophic information. Ye [13] defined the Hamming and Euclidean distances between the interval neutrosophic sets (INSs) and proposed the similarity measures between INSs based on the relationship between similarity measures and distances. Zhang et al. [14] defined the operations for INSs and put forward a comparison approach that was based on the related research of interval valued intuitionistic fuzzy sets (IVIFSs) and developed two interval neutrosophic number aggregation operators. Peng et al. [15] developed a new outranking approach for multi-criteria decision-making (MCDM) problems in the context of a simplified neutrosophic environment, where the truth-membership degree, indeterminacy-membership degree, and falsity-membership degree for each element are singleton subsets in [0,1] and defined some outranking relations for simplified neutrosophic number (SNNs) based on ELECTRE (ELimination and Choice Expressing REality), and the properties within the outranking relations are further discussed in detail. Zhang et al. [16] proposed a novel outranking approach for multi-criteria decision-making (MCDM) problems to address situations where there is a set of numbers in the real unit interval and not just a specific number with a neutrosophic set. Liu and Liu [17] proposed the neutrosophic number weighted power averaging (NNWPA) operator, the neutrosophic number weighted geometric power averaging (NNWGPA) operator, the generalized neutrosophic number weighted power averaging (GNNWPA) operator, and studied the properties of above operators are studied, such as idempotency, monotonicity, boundedness, and so on. Peng et al. [18] introduced the multi-valued neutrosophic sets (MVNSs), which allow for the truth-membership, indeterminacy- membership, and falsity-membership degree to have a set of crisp values between zero and one, respectively, and the multi-valued neutrosophic power weighted average (MVNPWA) operator and proposed the multi-valued neutrosophic power weighted geometric (MVNPWG) operator. Zhang et al. [19] presented a new correlation coefficient measure, which satisfies the requirement of this measure equaling one if and only if two interval neutrosophic sets (INSs) are the same and used the proposed weighted correlation coefficient measure of INSs to solve decision-making problems, which take into account the influence of the evaluations’ uncertainty and both the objective and subjective weights. Chen and Ye [20] presented the Dombi operations of single-valued neutrosophic numbers (SVNNs) based on the operations of the Dombi T-norm and T-conorm, and then proposed the single-valued neutrosophic Dombi weighted arithmetic average (SVNDWAA) operator and the single-valued neutrosophic Dombi weighted geometric average (SVNDWGA) operator to deal with the aggregation of SVNNs and investigated their properties. Liu and Wang [21] proposed the single-valued neutrosophic normalized weighted Bonferroni mean (SVNNWBM) operator on the basis of Bonferroni mean, the weighted Bonferroni mean (WBM), and the normalized WBM, and developed an approach to solve the multiple attribute decision-making problems with SVNNs that were based on the SVNNWBM operator. Wu et al. [22] defined the prioritized weighted average operator and the prioritized weighted geometric operator for simplified neutrosophic numbers (SNNs) and proposed two novel effective cross-entropy measures for SNSs. Li et al. [23] proposed the improved generalized weighted Heronian mean (IGWHM) operator and the
Symmetry 2018, 10, 131
3 of 36
improved generalized weighted geometric Heronian mean (IGWGHM) operator, the single valued neutrosophic number improved generalized weighted Heronian mean (NNIGWHM) operator, and single valued the neutrosophic number improved generalized weighted geometric Heronian mean (NNIGWGHM) operator for multiple attribute group decision making (MAGDM) problems, in which attribute values take the form of SVNNs. Wang et al. [24] combined the generalized weighted BM (GWBM) operator and generalized weighted geometric Bonferroni mean (GWGBM) operator with single valued neutrosophic numbers (SVNNs) to propose the generalized single-valued neutrosophic number weight BM (GSVNNWBM) operator and the generalized single-valued neutrosophic numbers weighted GBM (GSVNNWGBM) operator and developed the MADM methods based on these operators. Wei & Zhang [25] utilized power aggregation operators and the Bonferroni mean to develop some single-valued neutrosophic Bonferroni power aggregation operators and single-valued neutrosophic geometric Bonferroni power aggregation operators. Peng & Dai [26] initiated a new axiomatic definition of single-valued neutrosophic distance measure and similarity measure, which is expressed by a single-valued neutrosophic number that will reduce the information loss and retain more original information. Although SVNS theory has been successfully applied in some areas, the SVNS is also characterized by the truth-membership degree, indeterminacy-membership degree, and falsity-membership degree information. However, all of the above approaches are unsuitable to describe the truth-membership degree, indeterminacy-membership degree, and falsity-membership degree information of an element to a set by linguistic variables on the basis of the given linguistic term sets, which can reflect the decision maker’s confidence level when they are making an evaluation. In order to overcome this limit, we shall propose the concept of 2-tuple linguistic neutrosophic numbers set (2TLNNSs) to solve this problem based on the SVNS [4–6] and the 2-tuple linguistic information processing model [27–42]. Thus, how to aggregate these 2-tuple linguistic neutrosophic numbers is an interesting topic. To solve this issue, in this paper, we shall develop some 2-tuple linguistic neutrosophic information aggregation operators that are based on the traditional Bonferroni mean (BM) operations [43–50]. In order to do so, the remainder of this paper is set out as follows. In the next section, we shall propose the concept of 2TLNNSs. In Section 3, we shall propose some Bonferroni mean (BM) operators with 2TLNNs. In Section 4, we shall propose some generalized Bonferroni mean (GBM) operators with 2TLNNs. In Section 5, we shall propose some dual generalized Bonferroni mean (DGBM) operators with 2TLNNs. In Section 6, we shall present a numerical example for green supplier selection in order to illustrate the method that is proposed in this paper. Section 7 concludes the paper with some remarks. 2. Preliminaries In this section, we shall propose the concept of 2-tuple linguistic neutrosophic number sets (2TLNNSs) to solve this problem based on the SVNSs [6,7] and 2-tuple linguistic sets (2TLSs) [27,28]. 2.1. 2-Tuple Fuzzy Linguistic Representation Model Definition 1 ([27,28]). Let S = {si |i = 0, 1, . . . , t } be a linguistic term set with odd cardinality. Any label, si represents a possible value for a linguistic variable, and S can be defined as: ( S=
s0 = extremely poor, s1 = very poor, s2 = poor, s3 = medium, s4 = good, s5 = very good, s6 = extremely good.
)
Herrera and Martinez [27,28] developed the 2-tuple fuzzy linguistic representation model based on the concept of symbolic translation. It is used for representing the linguistic assessment information by means of a 2-tuple (si , ρi ), where si is a linguistic label for predefined linguistic term set S and ρi is the value of symbolic translation, and ρi ∈ [−0.5, 0.5).
Symmetry 2018, 10, 131
4 of 36
2.2. SVNSs Let X be a space of points (objects) with a generic element in fix set X, denoted by x. A single-valued neutrosophic sets (SVNSs) A in X is characterized as following [4–6]: A = {( x, TA ( x ), I A ( x ), FA ( x ))| x ∈ X }
(1)
where the truth-membership function TA ( x ), indeterminacy-membership I A ( x ), and falsitymembership function FA ( x ) are single subintervals/subsets in the real standard [0, 1], that is, TA ( x ) : X → [0, 1], I A ( x ) : X → [0, 1] and FA ( x ) : X → [0, 1] . The sum of TA ( x ), I A ( x ), and FA ( x ) satisfies the condition 0 ≤ TA ( x ) + I A ( x ) + FA ( x ) ≤ 3. Then, a simplification of A is denoted by A = {( x, TA ( x ), I A ( x ), FA ( x ))| x ∈ X }, which is a SVNS. For a SVNS {(x, TA (x), IA (x), FA (x))|x ∈ X }, the ordered triple components (TA (x), IA (x), FA (x)), are described as a single-valued neutrosophic number (SVNN), and each SVNN can be expressed as A = ( TA , I A , FA ), where TA ∈ [0, 1], I A ∈ [0, 1], FA ∈ [0, 1] TA ∈ [0, 1], I A ∈ [0, 1], FA ∈ [0, 1], and 0 ≤ TA + I A + FA ≤ 3. 2.3. 2TLNNSs Assume that L = {l0 , l1 , . . . , lt } is a 2TLSs with odd cardinality t + 1. If Definition 2. l = h(s T , a), (s I , b), (s F , c)i is defined for (s T , a), (s I , b), (s F , c) ∈ L and a, b, c ∈ [0, t], where (s T , a), (s I , b) and (s F , c) express independently the truth degree, indeterminacy degree, and falsity degree by 2TLSs, then 2TLNNSs is defined as follows: lj =
n
o s Tj , a j , s Ij , b j , s Fj , c j
(2)
where 0 ≤ ∆−1 s Tj , a j ≤ t, 0 ≤ ∆−1 s Ij , b j ≤ t, 0 ≤ ∆−1 s Fj , c j ≤ t, and 0 ≤ ∆−1 s Tj , a j + ∆−1 s Ij , b j + ∆−1 s Fj , c j ≤ 3t.
Definition 3. Let l1 = s T1 , a1 , s I1 , b1 , s F1 , c1 be a 2TLNN in L. Then the score and accuracy functions of l1 are defined as follows: ( S ( l1 ) = ∆
) 2t + ∆−1 s T1 , a1 − ∆−1 s I1 , b1 − ∆−1 s F1 , c1 , ∆−1 (S(l1 )) ∈ [0, t] 3
) t + ∆−1 s T1 , a1 − ∆−1 s F1 , c1 H ( l1 ) = ∆ , ∆−1 ( H (l1 )) ∈ [0, t]. 2
(3)
(
Definition 4. Let l1 = 2TLNNs, then (1) (2) (3) (4) (5)
if S(l1 ) if S(l1 ) if S(l1 ) if S(l1 ) if S(l1 )
(4)
s T1 , a1 , s I1 , b1 , s F1 , c1 and l2 = s T2 , a2 , s I2 , b2 , s F2 , c2 be two
< S(l2 ), then l1 < l2 ; > S(l2 ), then l1 > l2 ; = S(l2 ), H (l1 ) < H (l2 ), then l1 < l2 ; = S(l2 ), H (l1 ) > H (l2 ), then l1 > l2 ; = S(l2 ), H (l1 ) = H (l2 ), then l1 = l2 .
Definition 5. Let l1 = 2TLNNs, then
s T1 , a1 , s I1 , b1 , s F1 , c1 and l2 = s T2 , a2 , s I2 , b2 , s F2 , c2 be two
Symmetry 2018, 10, 131
(1)
(2)
(3)
(4)
5 of 36
−1 ∆ (s T1 ,a1 ) ∆−1 (s T1 ,a1 ) ∆−1 (s T2 ,a2 ) ∆−1 (s T2 ,a2 ) + − · , ∆ t t t t t −1 l1 ⊕ l2 = ; ∆ (s I1 ,b1 ) ∆−1 (s I2 ,b2 ) ∆−1 (s F1 ,c1 ) ∆−1 (s F2 ,c2 ) · , ∆ t · ∆ t t t t t −1 − 1 ∆ (s T1 ,a1 ) ∆ (s T2 ,a2 ) · , ∆ t t t − 1 − 1 − 1 − 1 ∆ (s I1 ,b1 ) ∆ (s I1 ,b1 ) ∆ (s I2 ,b2 ) ∆ (s I2 ,b2 ) ∆ t + − · , ; l1 ⊗ l2 = t t t t ∆−1 s ,c −1 s ,c −1 s ,c −1 s ,c ∆ ∆ ∆ ( ) ( ) ( ) ( ) F F 1 2 1 2 F F 1 2 1 2 ∆ t + − · t t t t !! ! λ λ ∆−1 (s T1 ,a1 ) ∆−1 (s I1 ,b1 ) ∆ t 1 − 1 − , ∆ t , t t ! λl1 = , λ > 0; λ −1 ∆ (s F1 ,c1 ) t ∆ t ! !! −1 λ λ ∆ (s T1 ,a1 ) ∆−1 (s I1 ,b1 ) ,∆ t 1− 1− , t t ∆ t λ !! , λ > 0. ( l1 ) = λ ∆−1 (s F1 ,c1 ) t ∆ t 1− 1−
2.4. BM Operators Definition 6 ([44]). Let p, q > 0 and bi (i = 1, 2, · · · , n) be a collection of nonnegative crisp numbers then the Bonferroni mean (BM) is defined as follows: !1/( p+q)
n
∑
BM p,q (b1 , b2 , · · · , bn ) =
i,j=1
p q bi b j
(5)
Then BM p,q is called Bonferroni mean (BM) operator. 3. 2TLNNWBM and 2TLNNWGBM Operators 3.1. 2TLNNWBM Operator To consider the attribute weights, the weighted Bonferroni mean (WBM) is defined, as follows. Definition 7 ([44]). Let p, q > 0 and bi (i = 1, 2, · · · , n) be a collection of nonnegative crisp numbers with n
the weights vector being ω = (ω1 , ω2 , · · · ωn ) T , thereby satisfying ωi ∈ [0, 1] and ∑ ωi = 1. The weighted i =1
Bonferroni mean (WBM) is defined as follows: p,q WBMω (b1 , b2 , · · ·
n
, bn ) =
∑
i,j=1
!1/( p+q) p q ω i ω j bi b j
(6)
Then we extend WBM to fuse the 2TLNNs and propose 2-tuple linguistic neutrosophic number weighted Bonferroni mean (2TLNNWBM) aggregation operator.
Definition 8. Let li = s Ti , ai , s Ii , bi , s Fi , ci be a set of 2TLNNs. The 2-tuple linguistic neutrosophic number weighted Bonferroni mean (2TLNNWBM) operator is: p,q 2TLNNWBMω (l1 , l2 , . . . , ln )
=
n
⊕ i,j=1
p ωi ω j li
q ⊗ lj
!1/( p+q) (7)
Symmetry 2018, 10, 131
6 of 36
Theorem 1. Let li = s Ti , ai , s Ii , bi , s Fi , ci be a set of 2TLNNs. The aggregated value by using 2TLNNWBM operators is also a 2TLNN where ! 1/( p+q) n p,q p q 2TLNNWBMω (l1 , l2 , . . . , ln ) = ⊕ ωi ω j li ⊗ l j i,j=1 !q !ωi ω j !1/( p+q) p n ∆−1 s Tj ,a j ∆−1 (s Ti ,ai ) , · ∆ t 1− ∏ 1− t t i,j=1 !q !ωi ω j !1/( p+q) p n ∆−1 s Ij ,b j ∆−1 (s Ii ,bi ) , = ∆ t 1 − 1 − ∏ 1 − 1 − · 1− t t i,j=1 !q !ωi ω j !1/( p+q) p n ∆−1 s Fj ,c j ∆−1 (s Fi ,ci ) · 1− 1− 1− t t ∆ t 1 − 1 − i,j∏ =1
(8)
Proof. According to Definition 5, we can obtain p li
p −1 ∆ (s Ti ,ai ) ,∆ t 1− 1− = ∆ t t (
q lj
=
∆ t
Thus,
!q ! ∆−1 s Tj ,a j t
,∆ t 1−
1−
∆−1 (s Ii ,bi ) t
p
,∆ t 1− 1−
!q !! ∆−1 s Ij ,b j t
,∆ t 1−
1−
∆−1 (s Fi ,ci ) t
p
!q !!) ∆−1 s Fj ,c j t
!q !! −1 p ∆−1 s Tj ,a j ∆ (s Ti ,ai ) · , ∆ t t t !q !! p ∆−1 s Ij ,b j ∆−1 (s Ii ,bi ) p q li ⊗ l j = · 1− , ∆ t 1− 1− t t ! !! q p ∆−1 s Fj ,c j ∆−1 (s Fi ,ci ) · 1− ∆ t 1− 1− t t
(9)
(10)
(11)
Thereafter, ωi ω j =
p
q
li ⊗ l j
!q !ωi ω j !! ∆−1 s Tj ,a j
∆ t 1− 1− · , t ! ! ! ω ω q i j p − 1 ∆ s Ij ,b j ∆−1 (s Ii ,bi ) ∆ t 1− 1− · 1 − , t t ! q ! ωi ω j ! p − 1 − 1 ∆ s Fj ,c j ∆ (s Fi ,ci ) · 1 − ∆ t 1− 1− t t
∆−1 (s Ti ,ai ) t
p
(12)
Furthermore, n p q ⊕ ωi ω j li ⊗ l j i,j=1 !q !ωi ω j !! −1 p n ∆−1 s Tj ,a j ∆ (s Ti ,ai ) ∆ t 1− ∏ 1− · , t t i,j=1 ! q ! ωi ω j ! p n ∆−1 s Ij ,b j ∆−1 (s Ii ,bi ) = ∆ t ∏ 1− 1− · 1− , t t i,j=1 ! ! ! q ωi ω j p n ∆−1 s Fj ,c j ∆−1 (s Fi ,ci ) · 1− ∆ t ∏ 1− 1− t t i,j=1
(13)
Symmetry 2018, 10, 131
7 of 36
Therefore, p,q 2TLNNWBMω (l1 , l2 , . . . , ln )
n
⊕
=
i,j=1
p ωi ω j li
q ⊗ lj
!1/( p+q)
!q !ωi ω j !1/( p+q) p −1 n ∆−1 s Tj ,a j ∆ (s Ti ,ai ) , · ∆t 1 − ∏ 1 − t t i,j=1 !q !ωi ω j !1/( p+q) p n ∆−1 s Ij ,b j ∆−1 (s Ii ,bi ) , = · 1− ∆ t 1 − 1 − ∏ 1 − 1 − t t i,j=1 !q !ωi ω j !1/( p+q) p n ∆−1 s Fj ,c j ∆−1 (s Fi ,ci ) · 1− t t ∆ t 1− 1− ∏ 1− 1− i,j=1
(14)
Hence, (8) is kept. Then, we need to prove that (8) is a 2TLNN. We need to prove two conditions, as follows: 1 2
0 ≤ ∆−1 (s T , a) ≤ t, 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t 0 ≤ ∆−1 (s T , a) + ∆−1 (s I , b) + ∆−1 (s F , c) ≤ 3t Let ∆−1 (s T ,a) t ∆−1 (s I ,b) t
n
1− ∏
=
1−
i,j=1 n
= 1− 1− ∏
∆−1 (s Ti ,ai ) t
1− 1−
i,j=1
∆−1 (s F ,c) t
n
= 1− 1− ∏
1 Since 0 ≤ Proof.
1− 1−
i,j=1
∆−1 (s Ti ,ai ) t
≤ 1, 0 ≤
·
!q !ωi ω j !1/( p+q) ∆−1 s Tj ,a j t
∆−1 (s Ii ,bi ) t
p
∆−1 (s Fi ,ci ) t
p
∆−1 s Tj ,a j t
∆−1 s Ti , ai t
0 ≤ 1−
p
0 ≤ 1− ∏
1−
i,j=1
0≤
n
1− ∏
i,j=1
1−
· 1−
t
!q !ωi ω j !1/( p+q) ∆−1 s Fj ,c j t
≤ 1 we get
q ! p −1 ∆ s Tj , a j ≤1 · t
Then, n
· 1−
!q !ωi ω j !1/( p+q) ∆−1 s Ij ,b j
∆−1 (s Ti ,ai ) t
∆−1 (s Ti ,ai ) t
p
·
p
·
! q ! ωi ω j ∆−1 s Tj ,a j t
(15)
≤1
!q !ωi ω j !1/( p+q) ∆−1 s Tj ,a j t
(16)
≤1
(17)
1 is maintained, similarly, we can get 0 ≤ ∆ −1 ( s I , b ) ≤ t That means 0 ≤ ∆−1 (s T , a) ≤ t, so − 1 − 1 2 Since 0 ≤ ∆ 0 ≤ ∆ (s F , c) ≤ t. (s T , a) ≤ t, 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t, we get − 1 − 1 − 1 0 ≤ ∆ (s T , a) + ∆ (s I , b) + ∆ (s F , c) ≤ 3t.
Example 1. Let h(s3 , 0.4), (s2 , −0.3), (s4 , 0.1)i, h(s2 , 0.3), (s1 , 0.2), (s4 , −0.1)i be two 2TLNNs, ( p, q) = (2, 3), ω = (0.4, 0.6) according to (8), we have
Symmetry 2018, 10, 131
(2,3) 2TLNNWBM(0.4,0.6)
8 of 36
! h(s3 , 0.4), (s2 , −0.3), (s4 , 0.1)i, h(s2 , 0.3), (s1 , 0.2), (s4 , −0.1)i !q !ωi ω j !1/( p+q) p −1 ∆−1 s Tj ,a j ∆ (s Ti ,ai ) , · 1− t t
n ∆t 1 − ∏ i,j=1 n = ∆ t 1 − 1 − ∏ 1 − 1 − i,j=1 n ∆ t 1− 1− ∏ 1− 1−
! ! ! 1/( p+q) ω ω q i j p ∆−1 s Ij ,b j ∆−1 (s Ii ,bi ) · 1 − , t t ! ! ! 1/ ( p + q ) ω ω q p i j − 1 − 1 ∆ s Fj ,c j ∆ (s Fi ,ci ) · 1 − t t
i,j=1
∆ 6 × 1 − = ∆ 6 × 1 − ∆ 6 × 1 −
0.4×0.6 2+1 3
× × 1− × 1− 0.6×0.4 3 0.6 × 0.6 2.3 2 3.4 2.3 2 2.3 3 × 1− 6 × 6 × 1− 6 × 6 1 2 3 0.4×0.4 2 3 0.4×0.6 2+3 1.7 1.7 1.7 1.2 1− 1− 1− 6 × 1− 6 × 1− 1− 6 × 1− 6 , 2 3 0.6×0.4 2 3 0.6×0.6 × 1 − 1 − 1.2 × 1 − 1.7 × 1 − 1 − 1.2 × 1 − 1.2 6 6 6 6 1 0.4 × 0.4 0.4 × 0.6 2+3 2 3 2 4.1 4.1 4.1 3.9 3 1− 1− 1− 6 × 1− 1− 6 × 1− 6 × 1− 6 0.6×0.4 3 0.6 × 0.6 3.9 2 3.9 2 4.1 3.9 3 × 1− 1− 6 × 1− 6 × 1− 1− 6 × 1− 6
3.4 6
2
3.4 6
3 0.4×0.4
3.4 6
2
2.3 3 6
,
= h(s3 , −0.173), (s1 , 0.384), (s4 , −0.024)i
Then, we will discuss some properties of 2TLNNWBM operator. Property 1. (Idempotency) If li =
s Ti , ai , s Ii , bi , s Fi , ci (i = 1, 2, . . . , n) are equal, then p,q
2TLNNWBMω (l1 , l2 , · · · , ln ) = l
(18)
Proof. Since li = l = h(s T , a), (s I , b), (s F , c)i, then p,q 2TLNNWBMω (l1 , l2 , . . . , ln )
n
!1/( p+q)
(l p
⊗ lq )
= ⊕ ωi ω j i,j=1 ! p −1 q ωi ω j 1/( p+q) n ∆−1 (s T ,a) ∆ (s T ,a) , ∆ t 1− ∏ 1− · t t i,j = 1 ! p q ωi ω j 1/( p+q) n − 1 − 1 ∆ (s I ,b) , = ∆ t 1 − 1 − ∏ 1 − 1 − · 1 − ∆ (ts I ,b) t i,j = 1 ! p q ωi ω j 1/( p+q) n − 1 − 1 ∆ (s F ,c) · 1 − ∆ (ts F ,c) ∆ t 1 − 1 − ∏ 1 − 1 − t i,j=1 1/( p+q) n n p + q ∑ ωi ∑ ω j −1 ∆ s ,a ( ) i =1 j =1 T 1 − 1 − ∆ t , t 1/( p+q) n n p + q ∑ ωi ∑ ω j − 1 ∆ s ,b ( ) i = 1 j = 1 I = ∆ t 1 − 1 − 1 − 1 − , t 1/ ( p + q ) n n ω ω ∑ ∑ i j p+q − 1 ∆ (s F ,c) i =1 j =1 ∆ t 1 − 1 − 1 − 1 − t = h(s T , a), (s I , b), (s F , c)i = l
Symmetry 2018, 10, 131
9 of 36
D E Property 2. (Monotonicity) Let lxi = s Tx , a xi , s Ix , bxi , s Fx , c xi (i = 1, 2, . . . , n) and i i i D E l yi = s Ty , ayi , s Iy , byi , s Fy , cyi (i = 1, 2, . . . , n) be two sets of 2TLNNs. If ∆−1 s Tx , a xi ≤ i i i i − 1 − 1 − 1 − 1 − 1 ∆ s Ty , ayi ,∆ s Ix , b x i ≥ ∆ s Iy , byi and ∆ s Fx , c xi ≥ ∆ s Fy , cyi hold for all i, then i
i
i
i
i
p,q
p,q
2TLNNWBMω (lx1 , lx2 , · · · , lxn ) ≤ 2TLNNWBMω ly1 , ly2 , · · · , lyn
(19)
D E p,q Let 2TLNNWBMω (lx1 , lx2 , · · · , lxn ) = s Tx , a xi , s Ix , bxi , s Fx , c xi (i = 1, 2, . . . , n) i i i D E p,q and 2TLNNWBMω ly1 , ly2 , · · · , lyn = s Ty , ayi , s Iy , byi , s Fy , cyi (i = 1, 2, . . . , n), given that i i i − 1 − 1 ∆ s Tx , a xi ≤ ∆ s Ty , ayi , we can obtain Proof.
i
i
q q p p −1 s , a ∆−1 s Tx , a x j ∆−1 s Ty , ay j ∆−1 s Tx , a xi ∆ y T y i j j i i · ≤ · t t t t
1 −
!p ∆−1 s Tx ,a xi t
i
q ωi ω j ∆−1 s Tx ,a x j
j
·
t
!p ∆−1 s Ty ,ayi
≥ 1 −
t
i
·
q ωi ω j ∆−1 s Ty ,ay j j
t
(20)
(21)
Thereafter, n
1 − ∏ 1 −
!p ∆−1 s Tx ,a xi t
i,j=1
i
·
q ωi ω j ∆−1 s Tx ,a x j j
t
n
!p ∆−1 s Ty ,ayi
≤ 1 − ∏ 1 −
t
i,j=1
i
·
q ωi ω j ∆−1 s Ty ,ay j j
t
(22)
Furthermore,
n 1 − ∏ 1 − i,j=1
n ≤ 1 − ∏ 1 − i,j=1
∆ −1
s Tx ,a xi t
!p
i
∆−1 s Ty ,ayi t
i
· !p
q ωi ω j 1/( p+q) ∆−1 s Tx ,a x j j
t
·
∆ −1
s Ty ,ay j
q ωi ω j 1/( p+q)
j
t
(23)
That means ∆−1 (s Tx , a x ) ≤ ∆−1 s Ty , ay . Similarly, we can obtain ∆−1 (s Ix , bx ) ≥ ∆−1 s Iy , by and ∆−1 (s Fx , c x ) ≥ ∆−1 s Fy , cy . If ∆−1 (s Tx , a x ) < ∆−1 s Ty , ay and ∆−1 (s Ix , bx ) ≥ ∆−1 s Iy , by and ∆−1 (s Fx , c x ) ≥ ∆−1 s Fy , cy p,q
p,q
2TLNNWBMω (lx1 , lx2 , · · · , lxn ) < G2TLNNWBMω ly1 , ly2 , · · · , lyn
If ∆−1 (s Tx , a x ) = ∆−1 s Ty , ay and ∆−1 (s Ix , bx ) > ∆−1 s Iy , by and ∆−1 (s Fx , c x ) > ∆−1 s Fy , cy p,q
p,q
2TLNNWBMω (lx1 , lx2 , · · · , lxn ) < G2TLNNWBMω ly1 , ly2 , · · · , lyn
If ∆−1 (s Tx , a x ) = ∆−1 s Ty , ay and ∆−1 (s Ix , bx ) = ∆−1 s Iy , by and ∆−1 (s Fx , c x ) = ∆−1 s Fy , cy p,q
p,q
2TLNNWBMω (lx1 , lx2 , · · · , lxn ) = 2TLNNWBMω ly1 , ly2 , · · · , lyn So, Property 2 is right.
Symmetry 2018, 10, 131
10 of 36
Property 3. (Boundedness) Let li = sTi , ai , s Ii , bi , sFi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs. If l + = maxi STi , ai , mini S Ii , bi , mini SFi , ci and l − = mini STi , ai , maxi S Ii , bi , maxi SFi , ci then p,q
l − ≤ 2TLNNWBMω (l1 , l2 , · · · , ln ) ≤ l + From Property 1,
(24)
p,q 2TLNNWBMω l1− , l2− , · · · , ln− = l − p,q 2TLNNWBMω l1+ , l2+ , · · · , ln+ = l +
From Property 2,
p,q
l − ≤ 2TLNNWBMω (l1 , l2 , · · · , ln ) ≤ l + 3.2. 2TLNNWGBM Operator Similarly to WBM, to consider the attribute weights, the weighted geometric Bonferroni mean (WGBM) is defined, as follows: Definition 9 ([51]). Let p, q > 0 and bi (i = 1, 2, · · · , n) be a collection of nonnegative crisp numbers with the n
weights vector being ω = (ω1 , ω2 , · · · ωn ) T , thereby satisfying ωi ∈ [0, 1] and ∑ ωi = 1. If i =1
p,q
WGBMω (b1 , b2 , · · · , bn ) =
n ω ω 1 pbi + qb j i j ∏ p + q i,j=1
(25)
Then we extend WGBM to fuse the 2TLNNs and propose 2-tuple linguistic neutrosophic number weighted geometric Bonferroni mean (2TLNNWGBM) operator.
Definition 10. Let li = s Ti , ai , s Ii , bi , s Fi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs with their weight vector be wi = (w1 , w2 , . . . , wn ) T , thereby satisfying wi ∈ [0, 1] and ∑in=1 wi = 1. If p,q
2TLNNWGBMω (l1 , l2 , . . . , ln ) =
ω ω n 1 ⊗ pli ⊕ ql j i j p + q i,j=1
(26)
p,q
Then we called 2TLNNWGBMω the 2-tuple linguistic neutrosophic number weighted geometric BM.
Theorem 2. Let li = s Ti , ai , s Ii , bi , s Fi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs. The aggregated value by using 2TLNNWGBM operators is also a 2TLNN where ω ω n p,q 1 2TLNNWGBMω (l1 , l2 , . . . , ln ) = p+ pli ⊕ ql j i j q ⊗ i,j=1 1 ! q ! ωi ω j ! p + p q − 1 − 1 n ∆ s Tj ,a j ∆ (s Ti ,ai ) · 1− ∆ t 1 − 1 − ∏ 1 − 1 − , t t i,j=1 1 ! q ! ωi ω j ! p + −1 p q − 1 n ∆ s Ij ,b j ∆ (s Ii ,bi ) = ∆t 1 − ∏ 1 − · , t t i,j=1 1 ! q ! ωi ω j ! p + −1 p q − 1 n ∆ s Fj ,c j ∆ (s Fi ,ci ) ∆t 1 − ∏ 1 − · t t i,j=1
(27)
Symmetry 2018, 10, 131
11 of 36
Proof. From Definition 5, we can obtain, ( ∆ t 1−
pli =
∆−1 s Ti , ai 1− t
( ql j =
∆ t 1−
1−
! p !!
∆−1 s Ii , bi ,∆ t t
!q !! ∆−1 s Tj ,a j t
,∆ t
!p!
!q ! ∆−1 s Ij ,b j t
∆−1 s Fi , ci ,∆ t t
,∆ t
! p !) (28)
!q !) ∆−1 s Fj ,c j
(29)
t
Thus, !q !! ∆−1 s Tj ,a j ∆−1 (s Ti ,ai ) p · 1− , t t ∆ t 1 − (1−) ! ! pli ⊕ ql j = q p p −1 −1 ∆−1 s Ij ,b j ∆ (s Ii ,bi ) ∆ (s Fi ,ci ) · · ∆ t , ∆ t t t t
!q ! ∆−1 s Fj ,c j t
(30)
Therefore, ω ω pli ⊕ ql j i j ! q ! ωi ω j ! p −1 s ,a −1 s ,a ∆ T ∆ j ( ) T i j i · 1− , ∆ t 1− 1− t t ! ! !! q ωi ω j −1 p ∆−1 s Ij ,b j ∆ (s Ii ,bi ) = ∆ t 1− 1− · , t t ! ! !! q ωi ω j −1 p ∆−1 s Fj ,c j ∆ (s Fi ,ci ) ∆ t 1 − 1 − · t t
(31)
Thereafter, ω ω n ⊗ pli ⊕ ql j i j i,j=1 ! q ! ωi ω j ! p n ∆−1 s Tj ,a j ∆−1 (s Ti ,ai ) ∆ t ∏ 1− 1− · 1− , t t i,j=1 ! ! !! q ωi ω j −1 p n ∆−1 s Ij ,b j ∆ (s Ii ,bi ) = ∆ t 1− ∏ 1− · , t t i,j=1 !q !ωi ω j !! −1 p n ∆−1 s Fj ,c j ∆ (s Fi ,ci ) · ∆ t 1− ∏ 1− t t i,j=1
(32)
Furthermore, ω ω n p,q 1 2TLNNWGBMω (l1 , l2 , . . . , ln ) = p+ pli ⊕ ql j i j q ⊗ i,j=1 1 ! q ! ωi ω j ! p + q p − 1 n ∆ s Tj ,a j ∆−1 (s Ti ,ai ) ∆ t 1 − 1 − ∏ 1 − 1 − · 1− , t t i,j = 1 1 ! q ! ωi ω j ! p + −1 p q − 1 n ∆ s Ij ,b j ∆ (s Ii ,bi ) = ∆t 1 − ∏ 1 − · , t t i,j = 1 1 ! q ! ωi ω j ! p + −1 p q − 1 n ∆ s Fj ,c j ∆ (s Fi ,ci ) ∆t 1 − ∏ 1 − · t t i,j=1
Hence, (27) is kept.
(33)
Symmetry 2018, 10, 131
12 of 36
Then, we need to prove that (27) is a 2TLNN. We need to prove two conditions, as follows: 1 2
0 ≤ ∆−1 (s T , a) ≤ t, 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t 0 ≤ ∆−1 (s T , a) + ∆−1 (s I , b) + ∆−1 (s F , c) ≤ 3t Let ∆ −1 ( s
T ,a )
= 1− 1− ∏
t ∆ −1 ( s t
n
1− 1−
∆ −1
(sTi ,ai )
I ,b )
1− ∏
=
1−
i,j=1
∆−1 (s Ii ,bi ) t
1 Since 0 ≤ Proof.
∆−1 (s Ti ,ai ) t
≤ 1, 0 ≤
0 ≤ 1−
1−
∆ −1
i,j=1
0 ≤ 1−
n
1− ∏
i,j=1
t
∆−1 s Ti , ai 1− t
n
∏ 1 −
∆ −1
∆−1 s Tj ,a j
Then, 0≤
p
n ∆ −1 ( s F , c ) = 1 − ∏ 1 − t i,j=1
1− 1−
· 1−
t
i,j=1
n
p
·
1 ! q ! ωi ω j ! p + q ∆−1 s Tj ,a j
t
1 ! q ! ωi ω j ! p + q ∆−1 s Ij ,b j
t
q ωi ω j p+1 q ! p −1 ∆ s Fj , c j s Fi , ci · t t
≤ 1 we get
q !p ∆−1 s Tj , a j ≤1 · 1 − t
(34)
q ωi ω j !p −1 s , a ∆ Tj j s Ti , ai · 1 − ≤1 t t
∆−1 (s Ti ,ai ) t
p
· 1−
1 ! q ! ωi ω j ! p + q ∆−1 s Tj ,a j
t
≤1
(35)
(36)
1 is maintained, similarly, we can get 0 ≤ ∆ −1 ( s I , b ) ≤ That means 0 ≤ ∆−1 (s T , a) ≤ t, so − 1 − 1 2 Since 0 ≤ ∆ t, 0 ≤ ∆ (s F , c) ≤ t. (s T , a) ≤ t, 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t, we get 0 ≤ ∆−1 (s T , a) + ∆−1 (s I , b) + ∆−1 (s F , c) ≤ 3t.
Symmetry 2018, 10, 131
13 of 36
Example 2. Let h(s3 , 0.4), (s2 , −0.3), (s4 , 0.1)i, h(s2 , 0.3), (s1 , 0.2), (s4 , −0.1)i be two 2TLNNs, ( p, q) = (2, 3), ω = (0.4, 0.6) according to (27), we have ! h(s3 , 0.4), (s2 , −0.3), (s4 , 0.1)i, h(s2 , 0.3), (s1 , 0.2), (s4 , −0.1)i 1 ! q ! ωi ω j ! p + p q −1 s ,a −1 s ,a n ∆ T ∆ j ( Ti i ) j · 1− ∆ t 1 − 1 − ∏ 1 − 1 − , t t
(2,3) 2TLNNWGBM(0.4,0.6)
i,j=1
n ∆t 1 − ∏
∆−1 (s Ii ,bi ) t
p
1 ! q ! ωi ω j ! p + q ∆−1 s Ij ,b j
1− · , t i,j=1 1 ! ! ! ω ω q −1 p i j p+q − 1 n ∆ s Fj ,c j ∆ (s Fi ,ci ) ∆ t 1 − 1 − · ∏ t t i,j=1 2 3 0.4×0.4 2 3 0.4×0.6 × 1 − 3.4 × 1 − 1 − 3.4 × 1 − 2.3 1 − 1 − 1 − 3.4 6 6 6 6 ∆ 3 0.6×0.4 6 × 1 − 0.6×0.6 2.3 2 3.4 2.3 3 2.3 2 × 1 − 1 − × 1 − × 1 − × 1 − 1 − 6 6 6 6 1 2 3 0.4×0.4 2 3 0.4×0.6 2+3 1.7 × 1.7 × 1 − 1.7 × 1.2 6 6 6 1− 6 1 − , = ∆ 6 × 0.6 × 0.4 0.6 × 0.6 2 3 2 3 × 1 − 1.2 × 1.7 × 1 − 1.2 × 1.2 6 6 6 6 2+1 3 2 3 0.4×0.4 2 0.4×0.6 4.1 4.1 4.1 3.9 3 × 1− 6 × 6 × 6 1− 6 1 − ∆ 6 × 0.6 × 0.4 2 4.1 3 2 3 0.6×0.6 × 1 − 3.9 × 6 × 1 − 3.9 × 3.9 6 6 6
=
2+1 3
,
= h(s3 , −0.334), (s1 , 0.434), (s4 , −0.018)i
Similar to 2TLNNWBM, the 2TLNNWGBM has the same properties, as follows. The proof are omitted here to save space. Property 4. (Idempotency) If li =
s Ti , ai , s Ii , bi , s Fi , ci (i = 1, 2, . . . , n) are equal, then p,q
2TLNNWGBMω (l1 , l2 , · · · , ln ) = l
(37)
D E Property 5. (Monotonicity) Let lxi = s Tx , a xi , s Ix , bxi , s Fx , c xi (i = 1, 2, . . . , n) and i i i D E l yi = s Ty , ayi , s Iy , byi , s Fy , cyi (i = 1, 2, . . . , n) be two sets of 2TLNNs. If ∆−1 s Tx , a xi ≤ i i i i ∆−1 s Ty , ayi ,∆−1 s Ix , bxi ≥ ∆−1 s Iy , byi and ∆−1 s Fx , c xi ≥ ∆−1 s Fy , cyi hold for all i, then i
i
i
i
p,q
i
p,q
2TLNNWGBMω (lx1 , lx2 , · · · , lxn ) ≤ 2TLNNWGBMω ly1 , ly2 , · · · , lyn
(38)
sTi , ai , s Ii , bi , sFi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs. If Property 6. (Boundedness) Let li = l + = maxi STi , ai , mini S Ii , bi , mini SFi , ci and l − = mini STi , ai , maxi S Ii , bi , maxi SFi , ci then p,q
l − ≤ 2TLNNWGBMω (l1 , l2 , · · · , ln ) ≤ l +
(39)
4. G2TLNNWBM and G2TLNNWGBM Operators 4.1. G2TLNNWBM Operator The primary advantage of BM is that it can determine the interrelationship between arguments. However, the traditional BM can only consider the correlations of any two aggregated arguments. Thereafter, Beliakov et al. [43] extended the BM and introduced the generalized BM(GBM) operator. Zhu et al. [51] introduced the generalized weighted BM(GWBM) operator, as follows.
Symmetry 2018, 10, 131
14 of 36
Definition 11 ([51]). Let p, q, r > 0 and bi (i = 1, 2, ¡ ¡ ¡ , n) be a collection of nonnegative crisp numbers with n
the weights vector being ω = (ω1 , ω2 , ¡ ¡ ¡ ωn ) T , thereby satisfying ωi ∈ [0, 1] and ∑ ωi = 1. The generalized i =1
weighted Bonferroni mean (GWBM) is defined as follows: p,q,r GWBMω (b1 , b2 , ¡ ¡ ¡
!1/( p+q+r)
n
∑
, bn ) =
i,j,k =1
p q ωi ω j ωk bi b j bkr
(40)
Then we extend GWBM to fuse the 2TLNNs and propose generalized 2-tuple linguistic neutrosophic number weighted Bonferroni mean (G2TLNNWBM) aggregation operator.
Definition 12. Let li = s Ti , ai , s Ii , bi , s Fi , ci be a set of 2TLNNs. The generalized 2-tuple linguistic neutrosophic number weighted Bonferroni mean (G2TLNNWBM) operator is: p,q,r G2TLNNWBMω (l1 , l2 , . . . , ln )
=
n
⊕ i,j,k =1
p ωi ω j ω k li
q ⊗ lj
⊗ lkr
!1/( p+q+r) (41)
Theorem 3. Let li = s Ti , ai , s Ii , bi , s Fi , ci be a set of 2TLNNs. The aggregated value by using G2TLNNWBM operators is also a 2TLNN where ! 1/( p+q+r) n p q p,q,r ⊕ ωi ω j ωk li ⊗ l j ⊗ lkr G2TLNNWBMω (l1 , l2 , . . . , ln ) = i,j,k =1   !q !r !ωi ω j ωk !1/( p+q+r )  p −1  n ∆−1 s Tj ,a j ∆−1 s T ,ak  ∆ (s Ti ,ai )  k ,  ∆ďŁt 1 − âˆ? 1 − ¡ ¡  t t t   i,j,k =1      ! !r !ωi ω j ωk !1/( p+q+r )    p  n ∆−1 s Ij ,b j ∆−1 s I ,bk ∆−1 (s Ii ,bi ) k  , = ∆ ďŁ t ďŁ1 − 1 − âˆ? 1 − 1 − ¡ 1 − ¡ 1 − t t t   i,j,k =1     !q  !r !ωi ω j ωk !1/( p+q+r )   p   n ∆−1 s Fj ,c j ∆−1 s F ,ck  ∆−1 (s Fi ,ci )  k ďŁ ďŁ ďŁ¸ďŁ¸  ¡ 1− ¡ 1−  t t t  ∆ t 1− 1− âˆ? 1− 1− i,j,k =1
             
(42)
            
Proof. According to Definition 5, we can obtain p
li =
p −1 ∆ (s Ti ,ai ) ,∆ t 1− 1− ∆ t t
( q lj
∆ t
= (
lkr
=
∆ t
!q ! ∆−1 s Tj ,a j t !r ! ∆−1 s T ,ak k
t
,∆ t 1−
,∆ t 1−
1−
1−
∆−1 (s Ii ,bi ) t
p
,∆ t 1− 1−
!q !! ∆−1 s Ij ,b j t
,∆ t 1−
!r !! ∆−1 s I ,bk k
t
,∆ t 1−
1−
1−
∆−1 (s Fi ,ci ) t
p
(43)
!q !!) ∆−1 s Fj ,c j
(44)
t !r !!) ∆−1 s F ,ck k
(45)
t
Thus,  !q !r !! −1 p ∆−1 s Tj ,a j  ∆−1 s T ,ak ∆ (s Ti ,ai )  k  ∆ t ¡ ¡ ,  t t t     ! !r !!  q p  −1 s ,b −1 s ,b ∆−1 s Ij ,b j ∆ ∆ I k ( ) p q I i k i li ⊗ l j ⊗ lkr = ¡ 1− ¡ 1− , ∆ t 1− 1− t t t    ! ! !!  r p   ∆−1 s Fj ,c j ∆−1 s F ,ck ∆−1 (s Fi ,ci )  k   ∆ t 1 − 1 − ¡ 1 − ¡ 1 −  t t t
                    
(46)
Symmetry 2018, 10, 131
15 of 36
Thereafter, p q ωi ω j ωk li ⊗ l j ⊗ lkr !q !r !ωi ω j ωk !! p −1 ∆−1 s Tj ,a j ∆−1 s T ,ak ∆ (s Ti ,ai ) k · · , ∆ t 1− 1− t t t ! ! ! q r ωi ω j ω k ! p ∆−1 s Ij ,b j ∆−1 s I ,bk ∆−1 (s Ii ,bi ) k = · 1− · 1− , ∆ t 1− 1− t t t !q ! r ! ωi ω j ω k ! p ∆−1 s Fj ,c j ∆−1 s F ,ck ∆−1 (s Fi ,ci ) k · 1− · 1− ∆ t 1− 1− t t t
(47)
Furthermore, n p q ⊕ ωi ω j ωk li ⊗ l j ⊗ lkr i,j,k =1 !q !r !ωi ω j ωk !! p −1 n ∆−1 s Tj ,a j ∆−1 s T ,ak ∆ (s Ti ,ai ) k · · , ∆ t 1− ∏ 1− t t t i,j,k =1 !q ! r ! ωi ω j ω k ! p n ∆−1 s Ij ,b j ∆−1 s I ,bk ∆−1 (s Ii ,bi ) k = ∆ t ∏ 1− 1− · 1− · 1− , t t t i,j,k =1 !q ! r ! ωi ω j ω k ! p n ∆−1 s Fj ,c j ∆−1 s F ,ck ∆−1 (s Fi ,ci ) k ∆ t ∏ 1− 1− · 1− · 1− t t t i,j,k =1
(48)
Therefore, n
p,q,r G2TLNNWBMω (l1 , l2 , . . . , ln )
p ωi ω j ω k li
q ⊗ lj
⊗ lkr
!1/( p+q+r)
= ⊕ i,j,k =1 !q !r !ωi ω j ωk !1/( p+q+r ) −1 p n ∆−1 s Tj ,a j ∆−1 s T ,ak ∆ (s Ti ,ai ) k , ∆t 1 − ∏ 1 − · · t t t i,j,k =1 !q !r !ωi ω j ωk !1/( p+q+r ) p n ∆−1 s Ij ,b j ∆−1 s I ,bk ∆−1 (s Ii ,bi ) k , = ∆ t 1− 1− ∏ 1− 1− · 1− · 1− t t t i,j,k =1 !q !r !ωi ω j ωk !1/( p+q+r ) p n ∆−1 s Fj ,c j ∆−1 s F ,ck ∆−1 (s Fi ,ci ) k · 1− · 1− t t t ∆ t 1− 1− ∏ 1− 1− i,j,k =1
(49)
Hence, (42) is kept. Then, we need to prove that (42) is a 2TLNN. We need to prove two conditions as follows: 1 2
0 ≤ ∆−1 (s T , a) ≤ t, 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t 0 ≤ ∆−1 (s T , a) + ∆−1 (s I , b) + ∆−1 (s F , c) ≤ 3t Let ∆−1 (s T ,a) t
=
n
1− ∏
1−
i,j,k =1 n
∆−1 (s I ,b) t
= 1− 1− ∏
∆−1 (s F ,c) t
= 1− 1− ∏
n
∆−1 (s Ti ,ai ) t
·
!q ∆−1 s Tj ,a j t
p
1− 1−
∆−1 (s Ii ,bi ) t
1− 1−
∆−1 (s Fi ,ci ) t
p
i,j,k =1
i,j,k =1
p
· 1− · 1−
!r !ωi ω j ωk !1/( p+q+r ) ∆−1 s T ,ak k
·
t
!q ∆−1 s Ij ,b j t !q ∆−1 s Fj ,c j t
· 1− · 1−
!r !ωi ω j ωk !1/( p+q+r ) ∆−1 s I ,bk k
t !r !ωi ω j ωk !1/( p+q+r ) ∆−1 s F ,ck k
t
Symmetry 2018, 10, 131
16 of 36
1 Since 0 ≤ Proof.
∆−1 (s Ti ,ai ) t
≤ 1, 0 ≤
∆−1 s Ti , ai t
0 ≤ 1−
∆−1 s Tj ,a j t
≤ 1, 0 ≤
∆−1 s T ,ak k
t
q ! p −1 ∆ s Tj , a j · · t
≤ 1 we get
∆−1 s Tk , ak t
!r
≤1
(50)
Then, q ! p −1 ∆ s , a T j s Ti , ai j · · t t
∆ −1
n
0 ≤ 1−
∏
i,j,k =1
0≤
1 −
n
1− ∏
1−
i,j,k =1
∆−1 (s Ti ,ai ) t
p
·
!q ∆−1 s Tj ,a j t
·
∆ −1
s Tk , ak t
!r
ωi ω j ω k
!r !ωi ω j ωk !1/( p+q+r ) ∆−1 s T ,ak k
t
≤1
≤1
(51)
(52)
1 That means 0 ≤ ∆−1 (s T , a) ≤ t, so is maintained, similarly, we can get 2 Since 0 ≤ ∆ −1 ( s T , a ) ≤ t, 0 ≤ 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t. ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t, we get 0 ≤ ∆−1 (s T , a) + ∆−1 (s I , b) + ∆−1 (s F , c) ≤ 3t.
Then we will discuss some properties of G2TLNNWBM operator. Property 7. (Idempotency) If li =
s Ti , ai , s Ii , bi , s Fi , ci (i = 1, 2, . . . , n) are equal, then p,q,r
G2TLNNWBMω (l1 , l2 , · · · , ln ) = l
(53)
Proof. Since li = l = h(s T , a), (s I , b), (s F , c)i, then p,q,r G2TLNNWBMω (l1 , l2 , . . . , ln )
n
p ωi ω j ω k li
q ⊗ lj
⊗ lkr
!1/( p+q+r)
= ⊕ i,j,k =1 !q !r !ωi ω j ωk !1/( p+q+r ) −1 p n ∆−1 s Tj ,a j ∆−1 s T ,ak ∆ (s Ti ,ai ) k , ∆ t 1− ∏ 1− · · t t t i,j,k =1 !q !r !ωi ω j ωk !1/( p+q+r ) p n ∆−1 s Ij ,b j ∆−1 s I ,bk ∆−1 (s Ii ,bi ) k , = ∆ t 1 − 1 − ∏ 1 − 1 − · 1− · 1− t t t i,j,k =1 !q !r !ωi ω j ωk !1/( p+q+r ) p −1 s ,c −1 s ,c −1 s ,c n ∆ ∆ F ∆ j F k ( ) F i j k i · 1− · 1− t t t ∆ t 1− 1− ∏ 1− 1− i,j,k =1
1/( p+q+r) n n n −1 p + q +r ∑ ωi ∑ ω j ∑ ω k ∆ (s T ,a) i =1 j =1 k =1 ∆ t 1 − 1 − , t 1/( p+q+r) ∑n ωi ∑n ω j ∑n ωk −1 (s ,b ) p +q +r ∆ i = 1 j = 1 k = 1 I = ∆ t 1 − 1 − 1 − 1 − , t 1/( p+q+r) n n n p + q +r ∑ ωi ∑ ω j ∑ ω k − 1 ∆ (s F ,c) i =1 j =1 k =1 ∆ t t 1 − 1 − 1 − 1 −
= h(s T , a), (s I , b), (s F , c)i = l
Symmetry 2018, 10, 131
17 of 36
D E Property 8. (Monotonicity) Let lxi = s Tx , a xi , s Ix , bxi , s Fx , c xi (i = 1, 2, . . . , n) and i i i D E l yi = s Ty , ayi , s Iy , byi , s Fy , cyi (i = 1, 2, . . . , n) be two sets of 2TLNNs. If ∆−1 s Tx , a xi ≤ i i i i − 1 − 1 − 1 − 1 − 1 ∆ s Ty , ayi ,∆ s Ix , b x i ≥ ∆ s Iy , byi and ∆ s Fx , c xi ≥ ∆ s Fy , cyi hold for all i, then i
i
i
i
i
p,q,r
p,q,r
G2TLNNWBMω (lx1 , lx2 , · · · , lxn ) ≤ G2TLNNWBMω
l y1 , l y2 , · · · , l y n
(54)
D E p,q,r Proof. Let G2TLNNWBMω (lx1 , lx2 , · · · , lxn ) = s Tx , a xi , s Ix , bxi , s Fx , c xi (i = 1, 2, . . . , n) D i i E i p,q,r and G2TLNNWBMω ly1 , ly2 , · · · , lyn = s Ty , ayi , s Iy , byi , s Fy , cyi (i = 1, 2, . . . , n), given i i i − 1 − 1 that ∆ s Tx , a xi ≤ ∆ s Ty , ayi , we can obtain i
i
∆ −1
s Tx ,a xi t
t
∆ −1
1 −
·
i
!p ∆−1 s Ty ,ayi
≤
!p
s Tx ,a xi t
i
!p
·
i
!p ∆−1 s Ty ,ayi
≥ 1 −
t
i
q ∆−1 s Tx ,a x j j
·
t
·
!r ∆−1 s Tx ,a xk t
q ∆−1 s Ty ,ay j j
q ∆−1 s Tx ,a x j j
·
t
t
q ∆−1 s Ty ,ay j j
k
(56)
! r ωi ω j ω k ∆−1 s Ty ,ayk
·
t
k
! r ωi ω j ω k ∆−1 s Tx ,a xk
·
t
(55)
!r ∆−1 s Ty ,ayk
·
t
k
t
k
Thereafter,
n
!p ∆−1 s Tx ,a xi
1 − ∏ 1 −
t
i,j,k =1
n
∆ −1
≤ 1 − ∏ 1 −
i
s Ty ,ayi t
i,j,k =1
·
q ∆−1 s Tx ,a x j j
·
i
·
t
!p
! r ωi ω j ω k ∆−1 s Tx ,a xk t
q ∆−1 s Ty ,ay j j
t
∆ −1
·
k
s Ty ,ayk t
! r ωi ω j ω k
k
(57)
Furthermore,
n
t 1 − ∏ 1 − i,j,k =1
n
≤ t 1 − ∏ 1 − i,j,k =1
!p ∆−1 s Tx ,a xi t
∆ −1
i
s Ty ,ayi t
i
·
!p
q ∆−1 s Tx ,a x j j
·
t
·
q ∆−1 s Ty ,ay j j
t
·
1/( p+q+r)
! r ωi ω j ω k ∆−1 s Tx ,a xk t
∆ −1
k
s Ty ,ayk t
k
! r ωi ω j ω k
1/( p+q+r)
(58)
That means ∆−1 (s Tx , a x ) ≤ ∆−1 s Ty , ay . Similarly, we can obtain ∆−1 (s Ix , bx ) ≥ ∆−1 s Iy , by and ∆−1 (s Fx , c x ) ≥ ∆−1 s Fy , cy . If ∆−1 (s Tx , a x ) < ∆−1 s Ty , ay and ∆−1 (s Ix , bx ) ≥ ∆−1 s Iy , by and ∆−1 (s Fx , c x ) ≥ ∆−1 s Fy , cy p,q,r
p,q,r
G2TLNNWBMω (lx1 , lx2 , · · · , lxn ) < G2LNNWBMω
l y1 , l y2 , · · · , l y n
Symmetry 2018, 10, 131
18 of 36
If ∆−1 (s Tx , a x ) = ∆−1 s Ty , ay and ∆−1 (s Ix , bx ) > ∆−1 s Iy , by and ∆−1 (s Fx , c x ) > ∆−1 s Fy , cy p,q,r
p,q,r
G2TLNNWBMω (lx1 , lx2 , · · · , lxn ) < DG2LNNWBMω
l y1 , l y2 , · · · , l y n
If ∆−1 (s Tx , a x ) = ∆−1 s Ty , ay and ∆−1 (s Ix , bx ) = ∆−1 s Iy , by and ∆−1 (s Fx , c x ) = ∆−1 s Fy , cy p,q,r
p,q,r
G2TLNNWBMω (lx1 , lx2 , · · · , lxn ) = G2TLNNWBMω
l y1 , l y2 , · · · , l y n
So Property 8 is right.
sTi , ai , s Ii , bi , sFi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs. If Property 9. (Boundedness) Let li = l + = maxi STi , ai , mini S Ii , bi , mini SFi , ci and l − = mini STi , ai , maxi S Ii , bi , maxi SFi , ci then p,q,r
l − ≤ G2TLNNWBMω (l1 , l2 , · · · , ln ) ≤ l + From Property 7,
(59)
p,q,r G2TLNNWBMω l1− , l2− , · · · , ln− = l − p,q,r G2TLNNWBMω l1+ , l2+ , · · · , ln+ = l +
From Property 8,
p,q,r
l − ≤ G2TLNNWBMω (l1 , l2 , · · · , ln ) ≤ l + 4.2. G2TLNNWGBM Operator Similarly to GWBM, to consider the attribute weights, the generalized weighted geometric Bonferroni mean (GWGBM) is defined, as follows. Definition 13 ([51]). Let p, q, r > 0 and bi (i = 1, 2, · · · , n) be a collection of nonnegative crisp numbers with n
the weights vector being ω = (ω1 , ω2 , · · · ωn ) T , thereby satisfying ωi ∈ [0, 1] and ∑ ωi = 1. If i =1
p,q,r
GWGBMω (b1 , b2 , · · · , bn ) =
n ω ω ω 1 pbi + qb j + rbk i j k ∏ p + q + r i,j,k=1
(60)
Then we extend GWGBM to fuse the 2TLNNs and propose generalized 2-tuple linguistic neutrosophic number weighted geometric Bonferroni mean (G2TLNNWGBM) aggregation operator.
Definition 14. Let li = s Ti , ai , s Ii , bi , s Fi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs with their weight vector be wi = (w1 , w2 , . . . , wn ) T , thereby satisfying wi ∈ [0, 1] and ∑in=1 wi = 1. If p,q,r
G2TLNNWGBMω (l1 , l2 , . . . , ln ) = p,q,r
Then we called G2TLNNWGBMω geometric BM.
1 p+q+r
n
⊗ i,j,k =1
pli ⊕ ql j ⊕ rlk
ωi ω j ω k
(61)
the generalized 2-tuple linguistic neutrosophic number weighted
Symmetry 2018, 10, 131
19 of 36
Theorem 4. Let li = s Ti , ai , s Ii , bi , s Fi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs. The aggregated value by using G2TLNNWGBM operators is also a 2TLNN where ω ω ω n p,q,r G2TLNNWGBMω (l1 , l2 , . . . , ln ) = p+1q+r ⊗ pli ⊕ ql j ⊕ rlk i j k i,j,k = 1 !q !r !ωi ω j ωk ! p+1q+r p n ∆−1 s Tj ,a j ∆−1 s T ,ak ∆−1 (s Ti ,ai ) k · 1− · 1− ∆ t 1 − 1 − ∏ 1 − 1 − , t t t i,j,k = 1 !q !r !ωi ω j ωk ! p+1q+r −1 p n ∆−1 s Ij ,b j ∆−1 s I ,bk ∆ (s Ii ,bi ) k ∆t 1 − ∏ 1 − · · = , t t t i,j,k =1 !q !r !ωi ω j ωk ! p+1q+r −1 p n ∆−1 s Fj ,c j ∆−1 s F ,ck ∆ (s Fi ,ci ) k ∆t 1 − ∏ 1 − · · t t t i,j,k =1
(62)
Proof. From Definition 5, we can obtain, ( ∆ t 1−
pli =
∆−1 s Ti , ai 1− t
( ql j =
∆ t 1−
( rlk =
∆ t 1−
1−
! p !!
∆−1 s Ii , bi ,∆ t t
!q !! ∆−1 s Tj ,a j t
∆−1 s Tk , ak 1− t
!r !!
,∆ t
!p!
!q ! ∆−1 s Ij ,b j t
∆−1 s Ik , bk ,∆ t t
∆−1 s Fi , ci ,∆ t t
,∆ t
!r !
! p !) (63)
!q !) ∆−1 s Fj ,c j t
∆−1 s Fk , ck ,∆ t t
(64)
!r !) (65)
Thus, pli ⊕ ql j ⊕ rlk !q p ∆−1 s Tj ,a j ∆−1 (s Ti ,ai ) · 1− · 1− t t ∆ t 1− 1− ! ! q r! −1 p ∆−1 s Ij ,b j ∆−1 s I ,bk ∆ (s Ii ,bi ) k = ∆ t · · , t t t !q !r ! −1 p ∆−1 s Fj ,c j ∆−1 s F ,ck ∆ (s Fi ,ci ) k · · ∆ t t t t
!r !! ∆−1 s T ,ak k
t
,
(66)
Therefore, ω ω ω pli ⊕ ql j ⊕ rlk i j k !q ! r ! ωi ω j ω k ! p ∆−1 s Tj ,a j ∆−1 s T ,ak ∆−1 (s Ti ,ai ) k · 1− · 1− , ∆ t 1− 1− t t t !q !r !ωi ω j ωk !! p −1 ∆−1 s Ij ,b j ∆−1 s I ,bk ∆ (s Ii ,bi ) k = · · , ∆ t 1− 1− t t t !q !r !ωi ω j ωk !! −1 p ∆−1 s Fj ,c j ∆−1 s F ,ck ∆ (s Fi ,ci ) k · · ∆ t 1− 1− t t t
(67)
Symmetry 2018, 10, 131
20 of 36
Thereafter, n
⊗ =
pli ⊕ ql j ⊕ rlk
i,j,k =1
ωi ω j ω k
n
∆ t ∏
1− 1−
i,j,k =1
∆−1 (s Ti ,ai ) t
n
∆ t 1− ∏
1−
i,j,k =1
n
∆ t 1− ∏
1−
· 1−
∆−1 (s Ii ,bi ) t
p
∆ −1
p
(s Fi ,ci )
t
∆−1 s Ij ,b j
·
!q
t
! r ! ωi ω j ω k ! ∆−1 s T ,ak k
· 1−
t
t !q ∆−1 s Fj ,c j
·
t
i,j,k =1
!q ∆−1 s Tj ,a j
p
· ·
∆−1 s I ,bk
!r !ωi ω j ωk !!
k
t ∆ −1
s F ,ck
, !r !ωi ω j ωk !!
,
(68)
k
t
Furthermore, ω ω ω n p,q,r G2TLNNWGBMω (l1 , l2 , . . . , ln ) = p+1q+r ⊗ pli ⊕ ql j ⊕ rlk i j k i,j,k = 1 !q !r !ωi ω j ωk ! p+1q+r p n ∆−1 s Tj ,a j ∆−1 s T ,ak ∆−1 (s Ti ,ai ) k · 1− · 1− ∆ t 1 − 1 − ∏ 1 − 1 − , t t t i,j,k = 1 !q !r !ωi ω j ωk ! p+1q+r −1 p n ∆−1 s Ij ,b j ∆−1 s I ,bk ∆ (s Ii ,bi ) k ∆t 1 − ∏ 1 − · = · , t t t i,j,k =1 !q !r !ωi ω j ωk ! p+1q+r −1 p n ∆−1 s Fj ,c j ∆−1 s F ,ck ∆ (s Fi ,ci ) k ∆t 1 − ∏ 1 − · · t t t i,j,k =1
(69)
Hence, (62) is kept. Then, we need to prove that (62) is a 2TLNN. We need to prove two conditions, as follows: 0 ≤ ∆−1 (s T , a) ≤ t, 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t 0 ≤ ∆−1 (s T , a) + ∆−1 (s I , b) + ∆−1 (s F , c) ≤ 3t
1 2
Let ∆−1 (s T ,a) t ∆−1 (s I ,b) t ∆−1 (s F ,c) t
n
= 1− 1− ∏
1− 1−
i,j,k =1
=
n
1− ∏
1−
i,j,k =1
=
n
1− ∏
1−
i,j,k =1
1 Since 0 ≤ Proof.
0 ≤ 1−
∆−1 (s Ti ,ai ) t
∆−1 (s Ti ,ai ) t
∆−1 (s Ii ,bi ) t
p
∆−1 (s Fi ,ci ) t
p
≤ 1, 0 ≤
∆−1 s Ti , ai 1− t
·
p
· 1−
!q ∆−1 s Ij ,b j t !q ∆−1 s Fj ,c j
·
t
∆−1 s Tj ,a j t
!q ∆−1 s Tj ,a j t
· ·
≤ 1, 0 ≤
· 1−
!r !ωi ω j ωk ! p+1q+r ∆−1 s T ,ak k
t
!r !ωi ω j ωk ! p+1q+r ∆−1 s I ,bk k
t !r !ωi ω j ωk ! p+1q+r ∆−1 s F ,ck k
t
∆−1 s T ,ak k
t
q !p ∆−1 s Tj , a j · · 1 − t
≤ 1 we get
∆−1 s Tk , ak 1− t
!r
≤1
(70)
Then, n
0≤
∏
i,j,k =1
1 −
∆−1 s Ti , ai 1− t
q !p ∆−1 s Tj , a j · · 1 − t
∆−1 s Tk , ak 1− t
!r
ωi ω j ω k
≤ 1 (71)
Symmetry 2018, 10, 131
0 ≤ 1−
21 of 36
n
1− ∏
1− 1−
i,j,k=1
∆−1 (s Ti ,ai ) t
p
· 1−
!q ∆−1 s Tj ,a j t
· 1−
!r !ωi ω j ωk ! p+1q+r ∆−1 s T ,ak k
t
≤1
(72)
1 is maintained, similarly, we can get 0 ≤ ∆ −1 ( s I , b ) ≤ That means 0 ≤ ∆−1 (s T , a) ≤ t, so − 1 − 1 2 Since 0 ≤ ∆ t, 0 ≤ ∆ (s F , c) ≤ t. (s T , a) ≤ t, 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t, we get − 1 − 1 − 1 0 ≤ ∆ (s T , a) + ∆ (s I , b) + ∆ (s F , c) ≤ 3t.
Similar to G2TLNNWBM, the G2TLNNWGBM has the same properties as follows. The proof are omitted here to save space. Property 10. (Idempotency) If li =
s Ti , ai , s Ii , bi , s Fi , ci (i = 1, 2, . . . , n) are equal, then p,q,r
G2TLNNWGBMω (l1 , l2 , · · · , ln ) = l
(73)
D
E Property 11. (Monotonicity) Let lxi = s Tx , a xi , s Ix , bxi , s Fx , c xi (i = 1, 2, . . . , n) and i i i D E l yi = s Ty , ayi , s Iy , byi , s Fy , cyi (i = 1, 2, . . . , n) be two sets of 2TLNNs. If ∆−1 s Tx , a xi ≤ i i i i − 1 − 1 − 1 − 1 − 1 ∆ s Ty , ayi ,∆ s Ix , b x i ≥ ∆ s Iy , byi and ∆ s Fx , c xi ≥ ∆ s Fy , cyi hold for all i, then i
i
i
i
i
p,q,r
p,q,r
G2TLNNWGBMω (lx1 , lx2 , · · · , lxn ) ≤ G2TLNNWGBMω
l y1 , l y2 , · · · , l y n
(74)
Property 12. (Boundedness) Let li = sTi , ai , s Ii , bi , sFi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs. If l + = maxi STi , ai , mini S Ii , bi , mini SFi , ci and l − = mini STi , ai , maxi S Ii , bi , maxi SFi , ci then p,q,r
l − ≤ G2TLNNWGBMω (l1 , l2 , · · · , ln ) ≤ l +
(75)
5. DG2TLNNWBM and DG2TLNNWGBM Operators 5.1. DG2TLNNWBM Operator However, the GBM still has some drawbacks, GBWM and GWGBM can only consider the interrelationship between any three aggregated arguments. So, Zhang et al. [52] introduced a new generalization of the traditional BM because the correlations are ubiquitous among all of the arguments. The new generalization of the traditional BM is called the dual GWBM (DGBM). The DGWBM is defined, as follows. Definition 15 ([52]). Let bi (i = 1, 2, · · · , n) be a collection of nonnegative crisp numbers with the weights n
vector being ω = (ω1 , ω2 , · · · ωn ) T , thereby satisfying ωi ∈ [0, 1] and ∑ ωi = 1. The dual generalized i =1
weighted Bonferroni mean (DGWBM) is defined as follows: R DGWBMw (b1 , b2 , . . . , bn )
n
=
∑
i1 ,i2 ,...,in =1
n
∏
j =1
rj wi j bij
!!1/∑n
j =1 r j
(76)
where R = (r1 , r2 , . . . , rn ) T is the parameter vector with ri ≥ 0 (i = 1, 2, . . . , n). Then we extend DGWBM to fuse the 2TLNNs and propose dual generalized 2-tuple linguistic neutrosophic number weighted Bonferroni mean (DG2TLNNWBM) operator.
Symmetry 2018, 10, 131
22 of 36
Definition 16. Let li = s Ti , ai , s Ii , bi , s Fi , ci be a set of 2TLNNs. The dual generalized 2-tuple linguistic neutrosophic number weighted Bonferroni mean (DG2TLNNWBM) operator is: R DG2TLNNWBMw ( l1 , l2 , . . . , l n )
n
=
n
⊕
⊗
i1 ,i2 ,...,in =1
j =1
rj wi j lij
!!1/∑n
i =1 r j
(77)
s Ti , ai , s Ii , bi , s Fi , ci be a set of 2TLNNs. The aggregated value by using Theorem 5. Let li = DG2TLNNWBM operators is also a 2TLNN where R DG2TLNNWBMw ( l1 , l2 , . . . , l n )
=
n
⊕
n
⊗
rj wi j li j
!!1/∑n
i =1 r j
i1 ,i2 ,...,in =1 j=1 r wi 1/∑in=1 r j j j −1 s ,a ∆ Ti ij n n j , ∆ ∏ 1 − ∏ 1 − 1 − t t 1 − i1 ,i2 ,...,in =1 j =1 r wi 1/∑in=1 r j j j − 1 s Ii ,bi ∆ n n j j , = ∆ t 1 − 1 − ∏ 1 − ∏ 1 − 1 − 1 − t i1 ,i2 ,...,in =1 j =1 r wi 1/∑in=1 r j j j −1 s ,c ∆ Fi i j n n j ∆ ∏ 1 − ∏ 1 − 1 − 1 − t t 1 − 1 − i1 ,i2 ,...,in =1 j =1
(78)
Proof. According to Definition 5, we can obtain rj li j
r j ∆−1 s Ti ,ai j j , ∆ t 1 − 1 − = ∆ t t
r j ∆−1 s Ii ,bi j
t
j
r j ∆−1 s Fi ,ci
, ∆ t 1 − 1 −
j
j
t
(79)
Thus, r wi j j ∆−1 s Ti ,ai j j ∆ t 1 − 1 − , ∆ t 1 − 1 − t rj r wi wi j li = j j j ∆−1 s Fi ,ci j j 1 − ∆ t 1 − t
r wi j j ∆−1 s Ii ,bi j
t
j
,
(80)
Thereafter, r wi j j ∆−1 s Ti ,ai n j j ∆ t ∏ 1 − 1 − , t j =1 r wi j j −1 s ,b ∆ I i n n ij j rj ⊗ wi j li = ∆ t 1 − ∏ 1 − 1 − 1 − , t j j =1 j =1 r wi j j − 1 ∆ s Fi ,ci n j j ∆ t 1 − ∏ 1 − 1 − 1 − t j = 1
(81)
Symmetry 2018, 10, 131
23 of 36
Furthermore, n
n
rj wi j li j
!
⊕ ⊗ j =1 r wi j j −1 s ,a ∆ T n n ij ij ∆ t 1 − ∏ , 1 − ∏ 1 − 1 − t i1 ,i2 ,...,in =1 j =1 r wi j j − 1 s Ii ,bi ∆ n n j j ∆t = ∏ , 1 − ∏ 1 − 1 − 1 − t i ,i ,...,i = 1 j = 1 n 1 2 r wi j j −1 s ,c ∆ F i n n ij j ∏ 1 − ∏ 1 − 1 − 1 − t ∆t i1 ,i2 ,...,in =1 j =1
i1 ,i2 ,...,in =1
(82)
Therefore, R DG2TLNNWBMw ( l1 , l2 , . . . , l n )
n
n
⊕
=
⊗
rj wi j li j
!!1/∑n
i =1 r j
i1 ,i2 ,...,in =1 j=1 r wi 1/∑in=1 r j j j −1 s ,a ∆ Ti ij n n j , ∆ ∏ 1 − ∏ 1 − 1 − t t 1 − i1 ,i2 ,...,in =1 j =1 r wi 1/∑in=1 r j j j − 1 ∆ s Ii ,bi n n j j , = ∆ t 1 − 1 − ∏ 1 − ∏ 1 − 1 − 1 − t j =1 i1 ,i2 ,...,in =1 r wi 1/∑in=1 r j j j −1 s ,c ∆ Fi i j n n j ∆ ∏ 1 − ∏ 1 − 1 − 1 − t t 1 − 1 − j =1 i1 ,i2 ,...,in =1
Hence, (78) is kept. Then, we need to prove that (78) is a 2TLNN. We need to prove two conditions as follows: 1 2
0 ≤ ∆−1 (s T , a) ≤ t, 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t 0 ≤ ∆−1 (s T , a) + ∆−1 (s I , b) + ∆−1 (s F , c) ≤ 3t Let ∆−1 (s T ,a) t
= 1 −
n
∏
i1 ,i2 ,...,in =1
= 1 − 1 −
= 1 − 1 −
n
∏
i1 ,i2 ,...,in =1
j
t
n 1 − ∏ 1 − 1 − 1 − j =1
n
∏
r wi 1/∑in=1 r j j j ∆−1 s Ti ,ai
j =1
∆−1 (s F ,c) t
1 − ∏ 1 − 1 −
∆−1 (s I ,b) t
n
i1 ,i2 ,...,in =1
n 1 − ∏ 1 − 1 − 1 − j =1
j
r wi 1/∑in=1 r j j j ∆−1 s Ii ,bi j
j
t
r wi 1/∑in=1 r j j j ∆−1 s Ii ,bi j
t
j
(83)
Symmetry 2018, 10, 131
1 Since 0 ≤ Proof.
24 of 36
∆−1 s Ti ,ai j
j
t
≤ 1, we get r j ∆−1 s Ti , ai j j 0 ≤ 1− ≤1 t
Then,
(84)
r j wi j −1 s , a ∆ n Ti ij j 0 ≤ 1 − ∏ 1 − 1 − ≤1 t j =1
(85)
Furthermore, r j wi j −1 s , a ∆ n Ti i j j ≤ 1 1 − ∏ 1 − 1 − t =1 j =1
n
∏
0 ≤ 1−
i1 ,i2 ,...,in
(86)
Therefore,
r j wi j 1/∑in=1 r j ∆−1 s Ti , ai j j ≤1 1 − ∏ 1 − 1 − t =1 j =1
n
0 ≤ 1 −
∏
i1 ,i2 ,...,in
n
(87)
1 is maintained, similarly, we can get 0 ≤ ∆ −1 ( s I , b ) ≤ That means 0 ≤ ∆−1 (s T , a) ≤ t, so 2 Since 0 ≤ ∆ −1 ( s T , a ) ≤ t, 0 ≤ ∆ −1 ( s I , b ) ≤ t, 0 ≤ ∆ −1 ( s F , c ) ≤ t, we get t, 0 ≤ ∆−1 (s F , c) ≤ t. − 1 − 1 0 ≤ ∆ (s T , a) + ∆ (s I , b) + ∆−1 (s F , c) ≤ 3t.
Then, we will discuss some properties of the DG2TLNNWBM operator. D E (Monotonicity) Let lxi = s Tx , a xi , s Ix , bxi , s Fx , c xi (i = 1, 2, . . . , n) and Property 13. i i i D E l yi = s Ty , ayi , s Iy , byi , s Fy , cyi (i = 1, 2, . . . , n) be two sets of 2TLNNs. If ∆−1 s Tx , a xi ≤ i i i i ∆−1 s Ty , ayi ,∆−1 s Ix , bxi ≥ ∆−1 s Iy , byi and ∆−1 s Fx , c xi ≥ ∆−1 s Fy , cyi hold for all i, then i
i
i
i
i
R R DG2TLNNWBMw l y1 , l y2 , · · · , l y n (lx1 , lx2 , · · · , lxn ) ≤ DG2TLNNWBMw
(88)
D E R Proof. Let DG2TLNNWBMw s Tx , a xi , s Ix , bxi , s Fx , c xi (i = 1, 2, . . . , n) ( l x1 , l x2 , · · · , l x n ) = D i i E i R and DG2TLNNWBMw l y1 , l y2 , · · · , l y n = s Ty , ayi , s Iy , byi , s Fy , cyi (i = 1, 2, . . . , n), given i i i − 1 − 1 that ∆ s Tx , a xi ≤ ∆ s Ty , ayi , we can obtain i
i
r j −1 s ∆ , a xi Txi j j ≤ t
1 −
∆ −1
r j wi
s Tx , a xi ij
t
j
j
r j −1 s ∆ , a yi Tyi j j t
≥ 1 −
∆ −1
(89)
r j wi
s Ty , ayi ij
t
j
j
(90)
Symmetry 2018, 10, 131
25 of 36
Thereafter,
! r j wi
n ∏ 1 − 1 −
∆−1 s Tx ,a xi ij
j
t
j =1
j
! r j wi
n ≤ ∏ 1 − 1 −
∆−1 s Ty ,ayi ij
j
t
j =1
j
(91)
Furthermore,
n
∏
i1 ,i2 ,...,in =1
! r j wi
n ∆ 1 − ∏ 1 − 1 −
−1
s Tx ,a xi ij
j
t
j =1
1 −
n
∏
i1 ,i2 ,...,in =1
≤ 1 −
≥
n
∏
i1 ,i2 ,...,in =1
n
∏
i1 ,i2 ,...,in =1
s Tx ,a xi
j
j
−1
s Ty ,ayi ij
t
j
s Ty ,ayi ij
t
j
j
(92)
1/∑in=1 r j
! r j wi
n 1 − ∏ 1 − 1 − j =1
ij
t
∆ −1
! r j wi
j =1
! r j wi ∆ −1
j =1
n ∆ 1 − ∏ 1 − 1 −
n 1 − ∏ 1 − 1 −
j
j
1/∑in=1 r j
That means ∆−1 (s Tx , a x ) ≤ ∆−1 s Ty , ay . Similarly, we can obtain ∆−1 (s Ix , bx ) ≥ ∆−1 s Iy , by and ∆−1 (s Fx , c x ) ≥ ∆−1 s Fy , cy . If ∆−1 (s Tx , a x ) < ∆−1 s Ty , ay and ∆−1 (s Ix , bx ) ≥ ∆−1 s Iy , by and ∆−1 (s Fx , c x ) ≥ ∆−1 s Fy , cy R R DG2TLNNWBMw l y1 , l y2 , · · · , l y n (lx1 , lx2 , · · · , lxn ) < DG2TLNNWBMw
If ∆−1 (s Tx , a x ) = ∆−1 s Ty , ay and ∆−1 (s Ix , bx ) > ∆−1 s Iy , by and ∆−1 (s Fx , c x ) > ∆−1 s Fy , cy R R DG2TLNNWBMw l y1 , l y2 , · · · , l y n (lx1 , lx2 , · · · , lxn ) < DG2TLNNWBMw
If ∆−1 (s Tx , a x ) = ∆−1 s Ty , ay and ∆−1 (s Ix , bx ) = ∆−1 s Iy , by and ∆−1 (s Fx , c x ) = ∆−1 s Fy , cy R R DG2TLNNWBMw l y1 , l y2 , · · · , l y n (lx1 , lx2 , · · · , lxn ) = DG2TLNNWBMw
So Property 13 is right.
Property 14. (Boundedness) Letli = sTi , ai , s Ii , bi , sFi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs. If l + = maxi STi , ai , mini S Ii , bi , mini SFi , ci and l − = mini STi , ai , maxi S Ii , bi , maxi SFi , ci then R l − ≤ DG2TLNNWBMw ( l1 , l2 , · · · , l n ) ≤ l +
(93)
Symmetry 2018, 10, 131
26 of 36
From Theorem 5, we can obtain R − − − DG2TLNNWBM w l1 , l2 , · · · , l n r wi 1/∑in=1 r j j j − 1 s Ti ,ai min∆ n n j j ∆t , ∏ 1 − 1 − ∏ 1 − 1 − t i1 ,i2 ,...,in =1 j =1 r wi 1/∑in=1 r j j j −1 s ,b max∆ Ii ij n n j , ∆ = ∏ 1 − ∏ 1 − 1 − 1 − t t 1 − 1 − i1 ,i2 ,...,in =1 j =1 r wi 1/∑in=1 r j j j − 1 s Fi ,ci max∆ n n j j ∆ t 1 − 1 − ∏ 1 − ∏ 1 − 1 − 1 − t i1 ,i2 ,...,in =1 j =1 R + + + DG2TLNNWBM w l1 , l2 , · · · , l n r wi 1/∑in=1 r j j j −1 s ,a max∆ Ti ij n n j , ∆ ∏ 1 − ∏ 1 − 1 − t t 1 − i ,i ,...,i = 1 j = 1 n 2 1 r wi 1/∑in=1 r j j j min∆−1 s Ii ,bi n n j j , ∆ t 1 − 1 − = ∏ 1 − ∏ 1 − 1 − 1 − t i1 ,i2 ,...,in =1 j =1 r wi 1/∑in=1 r j j j −1 s ,c min∆ Fi i j n n j ∆ ∏ 1 − ∏ 1 − 1 − 1 − t t 1 − 1 − j = 1 i ,i ,...,i = 1 n 2 1
From Property 13, R l − ≤ DG2TLNNWBMw ( l1 , l2 , · · · , l n ) ≤ l +
5.2. DG2TLNNWGBM Operator Similarly to DGWBM, to consider the attribute weights, the dual generalized weighted geometric Bonferroni mean (DGWGBM) is defined, as follows. Definition 17 ([52]). Let and bi (i = 1, 2, · · · , n) be a collection of nonnegative crisp numbers with the weights n
vector being ω = (ω1 , ω2 , · · · ωn ) T , thereby satisfying ωi ∈ [0, 1] and ∑ ωi = 1. If i =1
R DGWGBMw (b1 , b2 , · · ·
, bn ) =
n
1 ∑nj=1 r j
n
∏
i1 ,i2 ,...,in =1
∑
r j li j
!!∏n
j =1
wi
j
(94)
j =1
where R = (r1 , r2 , . . . , rn ) T is the parameter vector with ri ≥ 0 (i = 1, 2, . . . , n). Then we extend DGWGBM to fuse the 2TLNNs and propose dual generalized 2-tuple linguistic neutrosophic number weighted geometric Bonferroni mean (DG2TLNNWGBM) aggregation operator.
Definition 18. Let li = s Ti , ai , s Ii , bi , s Fi , ci (i = 1, 2, . . . , n)be a set of 2TLNNs with their weight vector be wi = (w1 , w2 , . . . , wn ) T , thereby satisfying wi ∈ [0, 1] and ∑in=1 wi = 1. If R DG2TLNNWGBMw ( l1 , l2 , . . . , l n ) =
1 ∑nj=1 r j
n
⊗ i1 ,i2 ,...,in =1
n
⊕ r j li j
j =1
!∏n
j =1
wi
j
(95)
R Then we called DG2TLNNWGBMw the dual generalized 2-tuple linguistic neutrosophic number weighted geometric BM.
Symmetry 2018, 10, 131
27 of 36
Theorem 6. Let li = s Ti , ai , s Ii , bi , s Fi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs. The aggregated value by using DG2TLNNWGBM operators is also a 2TLNN where R DG2TLNNWGBM w ( l1 , l2 , . . . , l n ) ! n ∏ j =1 wi j n n ⊗ ⊕ r j li j = ∑n 1 r j =1 j
i1 ,i2 ,...,in =1
j =1
1 r ∏ n wi ∑ n r j j =1 j =1 j j − 1 s Ti ,ai ∆ n n j j 1 − ∏ 1 − 1 − 1 − t ∆ , ∏ t i1 ,i2 ,...,in =1 j =1 1 r ∏ n wi ∑ n r j j = 1 j j = 1 j ∆−1 s Ii ,bi n n j j = 1 − ∆ t 1 − , ∏ ∏ t j =1 i1 ,i2 ,...,in =1 1 r ∏ n wi ∑ n r j j =1 j =1 j j − 1 s Fi ,ci ∆ n n j j 1 − ∏ t 1 − ∆ ∏ t i1 ,i2 ,...,in =1 j =1
(96)
Proof. From Definition 5, we can obtain, r j ∆−1 s Ti ,ai j j , ∆ t 1 − 1 − ∆ t 1 − 1 − t r r j li j = j −1 s ,c ∆ Fi i j j ∆t t
r j ∆−1 s Ti ,ai j
t
j
,
(97)
Thus, r r j j ∆−1 s Ti ,ai ∆−1 s Ii ,bi n n j j j j , ∆ t ∏ , ∆ t 1 − ∏ 1 − t t j =1 j =1 n r ⊕ r j li j = j j =1 ∆−1 s Fi ,ci n j j t ∆ t j∏ =1
(98)
Therefore, n
!∏n
j =1
wi
j
⊕ r j li j j =1 r ∏ n wi r ∏ n wi j j j =1 j =1 j j ∆−1 s Ti ,ai ∆−1 s Ii ,bi n n j j j j 1 − 1 − ∏ t 1 − , ∆ t 1 − ∆ ∏ , t t j =1 j =1 = r ∏ n wi j j =1 j ∆−1 s Fi ,ci n j j t ∆t1 − 1 − j∏ =1
(99)
Symmetry 2018, 10, 131
28 of 36
Thereafter,
n
n
⊗ i1 ,i2 ,...,in =1
⊕ r j li j
!∏n
j =1
wi
r ∏ n wi j j =1 j −1 s ,a ∆ T i n n ij j ∆t 1− ∏ 1− ∏ , t i1 ,i2 ,...,in =1 j =1 r ∏ n wi j j =1 j ∆−1 s Ii ,bi n n j j 1 − ∏ = ∆ t 1 − ∏ , t i ,i ,...,i = 1 j = 1 n 1 2 r ∏ n wi j j =1 j −1 s ,c ∆ F i n n ij j ∆ t 1 − 1− ∏ ∏ t i1 ,i2 ,...,in =1 j =1
j
j =1
(100)
Furthermore, R DG2TLNNWGBM w ( l1 , l2 , . . . , l n ) ! n ∏ j =1 wi j n n ⊗ ⊕ r j li j = ∑n 1 r j =1 j
i1 ,i2 ,...,in =1
j =1
1 r ∏ n wi ∑ n r j j =1 j j = 1 j ∆−1 s Ti ,ai n n j j 1 − ∏ 1 − ∆ t 1 − 1 − , ∏ t i ,i ,...,i = 1 j = 1 n 1 2 1 r ∏ n wi ∑ n r j j = 1 j j =1 j ∆−1 s Ii ,bi n n j j = 1− ∏ ∆ t 1 − , ∏ t j =1 i1 ,i2 ,...,in =1 1 r ∏ n wi ∑ n r j j =1 j j = 1 j ∆−1 s Fi ,ci n n j j 1 − ∏ 1 − ∆ t ∏ t i ,i ,...,i = 1 j = 1 n 1 2
Hence, (96) is kept. Then, we need to prove that (96) is a 2TLNN. We need to prove two conditions, as follows: 1 2
0 ≤ ∆−1 (s T , a) ≤ t, 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t 0 ≤ ∆−1 (s T , a) + ∆−1 (s I , b) + ∆−1 (s F , c) ≤ 3t Let ∆−1 (s T ,a) t
= 1 − 1 −
n
∏
i1 ,i2 ,...,in =1
∆−1 (s I ,b) t
= 1 −
n
∏
i1 ,i2 ,...,in =1
= 1 −
n
∏
n
i1 ,i2 ,...,in =1
1 − ∏ 1 −
n
∆ −1
s Ii ,bi j
n
1 − ∏ j =1
j
s Fi ,ci j
j
j
j
1
∏nj=1 wi j ∑nj=1 r j j
1
∏nj=1 wi j ∑nj=1 r j
r
t
j
r
t
∆ −1
s Ti ,ai
1
∏nj=1 wi j ∑nj=1 r j
r
t
j =1
∆ −1
j =1
1 − ∏
∆−1 (s F ,c) t
j
(101)
Symmetry 2018, 10, 131
1 Since 0 ≤ Proof.
29 of 36
∆−1 s Ti ,ai j
j
≤ 1, we get
t
r j ∆−1 s Ti , ai j j 0 ≤ 1 − ∏ 1 − ≤1 t j =1
n
Then,
r j ∏nj=1 wi j ∆−1 s Ti , ai j j ≤1 1 − ∏ 1 − t =1 j =1
n
0≤
∏
i1 ,i2 ,...,in
0 ≤ 1− 1 −
(102)
n
n
∏
i1 ,i2 ,...,in =1
n
1 − ∏ 1 −
∆ −1
j =1
(103)
r j ∏nj=1 wi j ∑nj=11 r j s Ti , ai j j ≤1 t
(104)
1 is maintained, similarly, we can get 0 ≤ ∆ −1 ( s I , b ) ≤ That means 0 ≤ ∆−1 (s T , a) ≤ t, so − 1 − 1 2 Since 0 ≤ ∆ t, 0 ≤ ∆ (s F , c) ≤ t. (s T , a) ≤ t, 0 ≤ ∆−1 (s I , b) ≤ t, 0 ≤ ∆−1 (s F , c) ≤ t, we get 0 ≤ ∆−1 (s T , a) + ∆−1 (s I , b) + ∆−1 (s F , c) ≤ 3t.
Similar to DG2TLNNWBM, the DG2TLNNWGBM has the same properties, as follows. The proof are omitted here to save space. D E (Monotonicity) Let lxi = s Tx , a xi , s Ix , bxi , s Fx , c xi (i = 1, 2, . . . , n) and Property 15. i i i D E l yi = s Ty , ayi , s Iy , byi , s Fy , cyi (i = 1, 2, . . . , n) be two sets of 2TLNNs. If ∆−1 s Tx , a xi ≤ i i i i ∆−1 s Ty , ayi ,∆−1 s Ix , bxi ≥ ∆−1 s Iy , byi and ∆−1 s Fx , c xi ≥ ∆−1 s Fy , cyi hold for all i, then i
i
i
i
i
R R DG2TLNNWGBMw l y1 , l y2 , · · · , l y n (lx1 , lx2 , · · · , lxn ) ≤ DG2TLNNWGBMw
(105)
Property 16. (Boundedness) Let li = sTi , ai , s Ii , bi , sFi , ci (i = 1, 2, . . . , n) be a set of 2TLNNs. If l + = maxi STi , ai , mini S Ii , bi , mini SFi , ci and l − = mini STi , ai , maxi S Ii , bi , maxi SFi , ci then R l − ≤ DG2TLNNWGBMw ( l1 , l2 , · · · , l n ) ≤ l +
(106)
6. Numerical Example and Comparative Analysis 6.1. Numerical Example Given the rise in environmental and resource conservation importance, green supply chain management has seen growth within industry. In addition, balancing economic development and environmental development is one of the critical issues faced by managers to help organizations maintain a strategically competitive position. Green supplier management as one of important part of green supply chain is also critical issue for effective green supply chain management. Thus, in this section, we shall present a numerical example to select green suppliers in green supply chain management with 2TLNNs in order to illustrate the method that is proposed in this paper. There is a panel with five possible green suppliers in green supply chain management Ai (i = 1, 2, 3, 4, 5) to 1 G1 is the select. The experts selects four attribute to evaluate the five possible green suppliers: 2 G2 is the environmental factors; 3 G3 is the delivery factor; and, 4 G4 product quality factor; is the price factor. The five possible green suppliers Ai (i = 1, 2, 3, 4, 5) are to be evaluated using
Symmetry 2018, 10, 131
30 of 36
the 2TLNNs by the three decision maker under the above four attributes (whose weighting vector ω = (0.35, 0.10, 0.25, 0.30), expert weighting vector, which are listed in Tables 1–3. Table 1. 2-tuple linguistic neutrosophic number (2TLNN) decision matrix ( R1 ).
A1 A2 A3 A4 A5
G1
G2
G3
G4
<(s3 , 0), (s2 , 0) (s1 , 0)> <(s4 , 0), (s3 , 0) (s2 , 0)> <(s4 , 0), (s3 , 0) (s4 , 0)> <(s5 , 0), (s5 , 0) (s4 , 0)> <(s3 , 0), (s4 , 0) (s2 , 0)>
<(s3 , 0), (s4 , 0) (s2 , 0)> <(s5 , 0), (s4 , 0) (s4 , 0)> <(s3 , 0), (s3 , 0) (s2 , 0)> <(s4 , 0), (s3 , 0) (s3 , 0)> <(s4 , 0), (s5 , 0) (s2 , 0)>
<(s4 , 0), (s4 , 0) (s2 , 0)> <(s4 , 0), (s4 , 0) (s2 , 0)> <(s5 , 0), (s3 , 0) (s2 , 0)> <(s5 , 0), (s4 , 0) (s5 , 0)> <(s3 , 0), (s4 , 0) (s1 , 0)>
<(s4 , 0), (s4 , 0) (s2 , 0)> <(s4 , 0), (s3 , 0) (s3 , 0)> <(s4 , 0), (s3 , 0) (s4 , 0)> <(s3 , 0), (s4 , 0) (s1 , 0)> <(s4 , 0), (s3 , 0) (s2 , 0)>
Table 2. 2TLNN decision matrix ( R2 ).
A1 A2 A3 A4 A5
G1
G2
G3
G4
<(s5 , 0), (s4 , 0) (s3 , 0)> <(s2 , 0), (s3 , 0) (s3 , 0)> <(s5 , 0), (s3 , 0) (s3 , 0)> <(s3 , 0), (s5 , 0) (s2 , 0)> <(s3 , 0), (s3 , 0) (s1 , 0)>
<(s3 , 0), (s5 , 0) (s2 , 0)> <(s3 , 0), (s3 , 0) (s3 , 0)> <(s3 , 0), (s2 , 0) (s2 , 0)> <(s4 , 0), (s2 , 0) (s3 , 0)> <(s3 , 0), (s4 , 0) (s5 , 0)>
<(s3 , 0), (s1 , 0) (s2 , 0)> <(s3 , 0), (s4 , 0) (s2 , 0)> <(s2 , 0), (s3 , 0) (s4 , 0)> <(s3 , 0), (s4 , 0) (s5 , 0)> <(s4 , 0), (s5 , 0) (s1 , 0)>
<(s4 , 0), (s1 , 0) (s3 , 0)> <(s5 , 0), (s4 , 0) (s3 , 0)> <(s3 , 0), (s2 , 0) (s4 , 0)> <(s5 , 0), (s1 , 0) (s4 , 0)> <(s5 , 0), (s3 , 0) (s2 , 0)>
Table 3. 2TLNN decision matrix ( R3 ).
A1 A2 A3 A4 A5
G1
G2
G3
G4
<(s5 , 0), (s3 , 0) (s1 , 0)> <(s4 , 0), (s2 , 0) (s2 , 0)> <(s2 , 0), (s1 , 0) (s4 , 0)> <(s5 , 0), (s4 , 0) (s4 , 0)> <(s3 , 0), (s3 , 0) (s2 , 0)>
<(s4 , 0), (s2 , 0) (s1 , 0)> <(s4 , 0), (s5 , 0) (s4 , 0)> <(s3 , 0), (s2 , 0) (s2 , 0)> <(s5 , 0), (s4 , 0) (s2 , 0)> <(s4 , 0), (s2 , 0) (s2 , 0)>
<(s4 , 0), (s4 , 0) (s3 , 0)> <(s3 , 0), (s2 , 0) (s3 , 0)> <(s4 , 0), (s5 , 0) (s2 , 0)> <(s3 , 0), (s4 , 0) (s5 , 0)> <(s4 , 0), (s2 , 0) (s3 , 0)>
<(s4 , 0), (s1 , 0) (s3 , 0)> <(s2 , 0), (s1 , 0) (s3 , 0)> <(s2 , 0), (s4 , 0) (s4 , 0)> <(s5 , 0), (s3 , 0) (s1 , 0)> <(s5 , 0), (s3 , 0) (s4 , 0)>
In the following, we utilize the approach that was developed to select green suppliers in green supply chain management. Definition 19. Let l j =
D
E s Tj , a j , s Ij , b j , s Fj , c j ( j = 1, 2, . . . , n) be a set of 2TLNNs with their weight
vector be wi = (w1 , w2 , . . . , wn ) T , thereby satisfying wi ∈ [0, 1] and ∑in=1 wi = 1, then we can obtain n
2TLNNWAA(l1 , l2 , . . . , ln ) = ∑ w j l j j =1 !w !! j −1 s ,a n n ∆ Tj j ∆ t 1 − 1 − ,∆ t∏ ∏ t j =1 j =1 !w ! = j −1 s ,c n ∆ F j j ∆ t∏ t
!w ! j ∆−1 s Ij ,b j t
j =1
(107)
j =1
n w 2TLNNWGA(l1 , l2 , . . . , ln ) = ∑ l j j j =1 !w ! j −1 s ,a n n ∆ T j j ,∆ t 1− ∏ 1− t ∆ t∏ j =1 j =1 !w !! = j − 1 n ∆ s Fj ,c j ∆ t 1− ∏ 1− t
,
!w !! j ∆−1 s Ij ,b j t
,
(108)
Step 1. According to 2TLNNs rij (i = 1, 2, 3, 4, 5, j = 1, 2, 3, 4), we can aggregate all of the 2TLNNs rij by using the 2TLNNWAA (2TLNNWGA) operator to get the overall 2TLNNs Ai (i = 1, 2, 3, 4, 5) of the green suppliers Ai . Then the aggregating results are shown in Table 4.
Symmetry 2018, 10, 131
31 of 36
Table 4. The aggregating results by the 2TLNNWAA operator.
A1 A2 A3 A4 A5 A1 A2 A3 A4 A5
G1
G2
<(s5 , 0.000), (s3 , 0.464), (s2 , −0.268)> <(s3 , 0.172), (s2 , 0.449), (s2 , 0.449)> <(s4 , 0.000), (s2 , −0.268), (s3 , 0.464)> <(s4 , 0.268), (s4 , 0.472), (s3 , −0.172)> <(s3 , 0.000), (s3 , 0.000), (s1 , 0.414)>
<(s4 , −0.449), (s3 , 0.162), (s1 , 0.414)> <(s4 , −0.449), (s4 , −0.127), (s3 , 0.464)> < (s3 , 0.000), (s2 , 0.000), (s2 , 0.000)> <(s5 , −0.414), (s3 , −0.172), (s2 , 0.449)> <(s4 , −0.449), (s3 , −0.172), (s3 , 0.162)>
G3
G4
<(s4 , −0.449), (s2 , 0.000), (s2 , 0.449)> <(s3 , 0.000), (s3 , −0.172), (s2 , 0.449)> <(s3 , 0.172), (s4 , −0.127), (s3 , 0.172)> <(s3 , 0.000), (s4 , 0.000), (s5 , 0.000)> <(s4 , 0.000), (s3 , 0.162), (s2 , −0.268)>
<(s4 , 0.000), (s1 , 0.000), (s3 , 0.000)> <(s4 , 0.000), (s2 , 0.000), (s3 , 0.000)> <(s3 , −0.464), (s3 , −0.172), (s4 , 0.000)> <(s5 , 0.000), (s2 , −0.268), (s2 , 0.000)> <(s5 , 0.000), (s3 , 0.000), (s3 , −0.172)>
Step 2. According to Table 4, we can aggregate all of the 2TLNNs rij by using the DG2TLNNWBM (DG2TLNNWGBM) operator to get the overall 2TLNNs Ai (i = 1, 2, 3, 4, 5) of the green suppliers Ai . Suppose that P = (1, 1, 1, 1), then the aggregating results are shown in Table 5. The aggregating Table 5. (DG2TLNNWGBM) operator.
A1 A2 A3 A4 A5
results
of
the
green
suppliers
by
the
DG2TLNNWBM
DG2TLNNWBM
DG2TLNNWGBM
<(s4 , 0.209), (s2 , 0.299), (s2 , 0.251)> <(s3 , 0.418), (s3 , −0.454), (s3 , −0.286)> <(s3 , 0.259), (s3 , −0.390), (s3 , 0.316)> <(s4 , 0.226), (s3 , 0.354), (s3 , 0.067)> <(s4 , −0.073), (s3 , 0.023), (s2 , 0.079)>
<(s4 , 0.192), (s2 , 0.336), (s2 , 0.262)> <(s3 , 0.414), (s3 , −0.446), (s3 , −0.283)> <(s3 , 0.250), (s3 , −0.371), (s3 , 0.326)> <(s4 , 0.201), (s3 , 0.396), (s3 , 0.108)> <(s4 , −0.098), (s3 , 0.023), (s2 , 0.095)>
Step 3. According to the aggregating results shown in Table 5 and the score functions of the green suppliers are shown in Table 6. Table 6. The score functions of the green suppliers.
A1 A2 A3 A4 A5
DG2TLNNWBM
DG2TLNNWGBM
(s4 , −0.114) (s3 , 0.386) (s3 , 0.111) (s3 , 0.268) (s4 , −0.392)
(s4 , −0.135) (s3 , 0.381) (s3 , 0.098) (s3 , 0.232) (s4 , −0.405)
Step 4. According to the score functions shown in Table 6 and the comparison formula of score functions, the ordering of the green suppliers is shown in Table 7. Note that “>” means “preferred to”. As we can see, depending on the aggregation operators that were used, the best green supplier is A1 . Table 7. Ordering of the green suppliers. Ordering DG2TLNNWBM DG2TLNNWGBM
A1 > A5 > A2 > A4 > A3 A1 > A5 > A2 > A4 > A3
Symmetry 2018, 10, 131
32 of 36
6.2. Influence of the Parameter on the Final Result In order to show the effects on the ranking results by changing parameters of P in the DG2TLNNWBM (DG2TLNNWGBM) operators, all of the results are shown in Tables 8 and 9. Table 8. Ranking results for different operational parameters of the DG2TLNNWBM operator. P
s(A1 )
s(A2 )
s(A3 )
s(A4 )
s(A5 )
(1, 1, 1, 1) (2, 2, 2, 2) (3, 3, 3, 3) (4, 4, 4, 4) (5, 5, 5, 5) (6, 6, 6, 6) (7, 7, 7, 7) (8, 8, 8, 8) (9, 9, 9, 9) (10, 10, 10, 10)
(s4 , −0.114) (s5 , 0.085) (s5 , 0.333) (s5 , 0.410) (s5 , 0.445) (s5 , 0.466) (s5 , 0.482) (s5 , 0.496) (s6 , −0.492) (s5 , −0.482)
(s3 , 0.386) (s5 , −0.267) (s5 , 0.032) (s5 , 0.121) (s5 , 0.156) (s5 , 0.174) (s5 , 0.185) (s5 , 0.195) (s5 , 0.203) (s5 , 0.210)
(s3 , 0.111) (s5 , −0.455) (s5 , −0.078) (s5 , 0.055) (s5 , 0.115) (s5 , 0.149) (s5 , 0.171) (s5 , 0.187) (s5 , 0.200) (s5 , 0.210)
(s3 , 0.268) (s5 , −0.263) (s5 , 0.163) (s5 , −0.326) (s5 , 0.404) (s5 , 0.448) (s5 , 0.476) (s5 , 0.495) (s6 , −0.490) (s6 , −0.479)
(s4 , −0.392) (s5 , −0.073) (s5 , 0.234) (s5 , 0.339) (s5 , 0.389) (s5 , 0.421) (s5 , 0.444) (s5 , 0.463) (s5 , 0.479) (s5 , 0.493)
Ordering A1 A1 A1 A1 A1 A1 A1 A1 A4 A4
> A5 > A5 > A5 > A5 > A4 > A4 > A4 > A4 > A1 > A1
> A2 > A4 > A4 > A4 > A5 > A5 > A5 > A5 > A5 > A5
> A4 > A2 > A2 > A2 > A2 > A2 > A2 > A2 > A2 > A2
> A3 > A3 > A3 > A3 > A3 > A3 > A3 > A3 > A3 > A3
Table 9. Ranking results for different operational parameters of the DG2TLNNWGBM operator. P
s(A1 )
s(A2 )
s(A3 )
s(A4 )
s(A5 )
(1, 1, 1, 1) (2, 2, 2, 2) (3, 3, 3, 3) (4, 4, 4, 4) (5, 5, 5, 5) (6, 6, 6, 6) (7, 7, 7, 7) (8, 8, 8, 8) (9, 9, 9, 9) (10, 10, 10, 10)
(s4 , −0.135) (s3 , −0.046) (s3 , −0.430) (s2 , 0.380) (s2 , 0.271) (s2 , 0.200) (s2 , 0.150) (s2 , 0.113) (s2 , 0.085) (s2 , 0.063)
(s3 , 0.381) (s3 , −0.422) (s2 , 0.326) (s2 , 0.218) (s2 , 0.156) (s2 , 0.111) (s2 , 0.073) (s2 , 0.040) (s2 , 0.011) (s2 , −0.016)
(s3 , 0.098) (s2 , 0.283) (s2 , 0.018) (s2 , −0.102) (s2 , −0.174) (s2 , −0.225) (s2 , −0.263) (s2 , −0.294) (s2 , −0.320) (s2 , −0.343)
(s3 , 0.232) (s2 , 0.300) (s2 , −0.122) (s2 , −0.352) (s2 , −0.498) (s1 , 0.400) (s1 , 0.326) (s1 , 0.269) (s1 , 0.225) (s1 , 0.189)
(s4 , −0.405) (s3 , −0.242) (s2 , 0.450) (s2 , 0.310) (s2 , 0.233) (s2 , 0.186) (s2 , 0.153) (s2 , 0.130) (s2 , 0.113) (s2 , 0.099)
Ordering A1 A1 A1 A1 A1 A1 A5 A5 A5 A5
> A5 > A5 > A5 > A5 > A5 > A5 > A1 > A1 > A1 > A1
> A2 > A2 > A2 > A2 > A2 > A2 > A2 > A2 > A2 > A2
> A4 > A4 > A3 > A3 > A3 > A3 > A3 > A3 > A3 > A3
> A3 > A3 > A4 > A4 > A4 > A4 > A4 > A4 > A4 > A4
6.3. Comparative Analysis Then, we compare our proposed method with other existing methods, including the LNNWAA operator and the LNNWGA operator proposed by Fang & Ye [53] and cosine measures of linguistic neutrosophic numbers [54]. The comparative results are shown in Table 10. Table 10. Ordering of the green suppliers. Ordering LNNWAA [53] LNNWGA [53] C w1 LNNs [54] C w2 LNNs [54]
A1 A1 A1 A1
> A5 > A5 > A5 > A5
> A4 > A4 > A2 > A2
> A2 > A2 > A3 > A4
> A3 > A3 > A4 > A3
From above, we can that we get the same results to show the practicality and effectiveness of the proposed approaches. However, the existing aggregation operators, such as the LNNWAA operator and the LNNWGA operator, do not consider the information about the relationship between the arguments being aggregated, and thus cannot eliminate the influence of unfair arguments on the decision result. Our proposed DG2TLNNWBM and DG2TLNNWGBM operators consider the information about the relationship among arguments being aggregated. 7. Conclusions In this paper, we investigate the MADM problems with 2TLNNs. Then, we utilize the Bonferroni mean (BM) operator, generalized Bonferroni mean (GBM) operator, and dual generalized Bonferroni
Symmetry 2018, 10, 131
33 of 36
mean (DGBM) operator to develop some Bonferroni mean aggregation operators with 2TLNNs: 2-tuple linguistic neutrosophic number weighted Bonferroni mean (2TLNNWBM) operator, 2-tuple linguistic neutrosophic number weighted geometric Bonferroni mean (2TLNNWGBM) operator, generalized 2-tuple linguistic neutrosophic number weighted Bonferroni mean (G2TLNNWBM) operator, generalized 2-tuple linguistic neutrosophic number weighted geometric Bonferroni mean (G2TLNNWGBM) operator, dual generalized 2-tuple linguistic neutrosophic number weighted Bonferroni mean (DG2TLNNWBM) operator, and dual generalized 2-tuple linguistic neutrosophic number weighted geometric Bonferroni mean (DG2TLNNWGBM) operator. The prominent characteristic of these proposed operators are studied. Then, we have utilized these operators to develop some approaches to solve the MADM problems with 2TLNNs. Finally, a practical example for green supplier selection in green supply chain management is given to verify the developed approach and to demonstrate its practicality and effectiveness. In the future, the application of the proposed aggregating operators of 2TLNNs needs to be explored in the decision making, risk analysis, and many other fields under uncertain environments [55–77]. Acknowledgments: The work was supported by the National Natural Science Foundation of China under Grant No. 71571128 and the Humanities and Social Sciences Foundation of Ministry of Education of the People’s Republic of China (16XJA630005) and the Construction Plan of Scientific Research Innovation Team for Colleges and Universities in Sichuan Province (15TD0004). Author Contributions: Jie Wang, Guiwu Wei and Yu Wei conceived and worked together to achieve this work, Jie Wang compiled the computing program by Matlab and analyzed the data, Jie Wang and Guiwu Wei wrote the paper. Finally, all the authors have read and approved the final manuscript. Conflicts of Interest: The authors declare no conflicts of interest.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Zadeh, L.A. Fuzzy sets. Inf. Control 1965, 8, 338–353. [CrossRef] Atanassov, K. Intuitionistic fuzzy sets. Fuzzy Sets Syst. 1986, 20, 87–96. [CrossRef] Atanassov, K.; Gargov, G. Interval-valued intuitionistic fuzzy sets. Fuzzy Sets Syst. 1989, 31, 343–349. [CrossRef] Smarandache, F. Neutrosophy: Neutrosophic Probability, Set, and Logic: Analytic Synthesis & Synthetic Analysis; American Research Press: Rehoboth, DE, USA, 1998. Wang, H.; Smarandache, F.; Zhang, Y.Q.; Sunderraman, R. Single valued neutrosophic sets. Multispace Multistructure 2010, 4, 410–413. Wang, H.; Smarandache, F.; Zhang, Y.Q.; Sunderraman, R. Interval Neutrosophic Sets and Logic: Theory and Applications in Computing; Hexis: Phoenix, AZ, USA, 2005. Ye, J. A multicriteria decision-making method using aggregation operators for simplified neutrosophic sets. J. Intell. Fuzzy Syst. 2014, 26, 2459–2466. Ye, J. Multicriteria decision-making method using the correlation coefficient under single-valued neutrosophic environment. Int. J. Gen. Syst. 2013, 42, 386–394. [CrossRef] Broumi, S.; Smarandache, F. Correlation coefficient of interval neutrosophic set. Appl. Mech. Mater. 2013, 436, 511–517. [CrossRef] Biswas, P.; Pramanik, S.; Giri, B.C. TOPSIS method for multi-attribute group decision-making under single-valued neutrosophic environment. Neural Comput. Appl. 2016, 27, 727–737. [CrossRef] Liu, P.D.; Chu, Y.C.; Li, Y.W.; Chen, Y.B. Some generalized neutrosophic number Hamacher aggregation operators and their application to Group Decision Making. Int. J. Fuzzy Syst. 2014, 16, 242–255. Şahin, R.; Liu, P.D. Maximizing deviation method for neutrosophic multiple attribute decision making with incomplete weight information. Neural Comput. Appl. 2016, 27, 2017–2029. [CrossRef] Ye, J. Similarity measures between interval neutrosophic sets and their applications in multicriteria decision-making. J. Intell. Fuzzy Syst. 2014, 26, 165–172. Zhang, H.Y.; Wang, J.Q.; Chen, X.H. Interval neutrosophic sets and their application in multicriteria decision making problems. Sci. Word J. 2014, 2014, 1–15. [CrossRef] [PubMed]
Symmetry 2018, 10, 131
15. 16. 17. 18.
19.
20. 21. 22.
23. 24. 25. 26. 27. 28.
29.
30. 31. 32. 33. 34. 35.
36. 37.
34 of 36
Peng, J.J.; Wang, J.Q.; Zhang, H.Y.; Chen, X.H. An outranking approach for multi-criteria decision-making problems with simplified neutrosophic sets. Appl. Soft Comput. 2014, 25, 336–346. [CrossRef] Zhang, H.; Wang, J.Q.; Chen, X.H. An outranking approach for multi-criteria decision-making problems with interval-valued neutrosophic sets. Neural Comput. Appl. 2016, 27, 615–627. [CrossRef] Liu, P.D.; Xi, L. The neutrosophic number generalized weighted power averaging operator and its application in multiple attribute group decision making. Int. J. Mach. Learn. Cybernet. 2016. [CrossRef] Peng, J.J.; Wang, J.Q.; Wu, X.H.; Wang, J.; Chen, X.H. Multi-valued neutrosophic sets and power aggregation operators with their applications in multi-criteria group decision-making problems. Int. J. Comput. Intell. Syst. 2015, 8, 345–363. [CrossRef] Zhang, H.Y.; Ji, P.; Wang, J.Q.; Chen, X.H. An improved weighted correlation coefficient based on integrated weight for interval neutrosophic sets and its application in multi-criteria decision-making problems. Int. J. Comput. Intell. Syst. 2015, 8, 1027–1043. [CrossRef] Chen, J.Q.; Ye, J. Some Single-Valued Neutrosophic Dombi Weighted Aggregation Operators for Multiple Attribute Decision-Making. Symmetry 2017, 9, 82. [CrossRef] Liu, P.D.; Wang, Y.M. Multiple attribute decision making method based on single-valued neutrosophic normalized weighted Bonferroni mean. Neural Comput. Appl. 2014, 25, 2001–2010. [CrossRef] Wu, X.H.; Wang, J.Q.; Peng, J.J.; Chen, X.H. Cross-entropy and prioritized aggregation operator with simplified neutrosophic sets and their application in multi-criteria decision-making problems. J. Intell. Fuzzy Syst. 2016, 18, 1104–1116. [CrossRef] Li, Y.; Liu, P.; Chen, Y. Some Single Valued Neutrosophic Number Heronian Mean Operators and Their Application in Multiple Attribute Group Decision Making. Informatica 2016, 27, 85–110. [CrossRef] Wang, J.; Tang, X.; Wei, G. Models for Multiple Attribute Decision-Making with Dual Generalized Single-Valued Neutrosophic Bonferroni Mean Operators. Algorithms 2018, 11, 2. [CrossRef] Wei, G.W.; Zhang, Z.P. Some Single-Valued Neutrosophic Bonferroni Power Aggregation Operators in Multiple Attribute Decision Making. J. Ambient Intell. Humaniz. Comput. 2018. [CrossRef] Peng, X.D.; Dai, J.G. Approaches to single-valued neutrosophic MADM based on MABAC, TOPSIS and new similarity measure with score function. Neural Comput. Appl. 2018, 29, 939–954. [CrossRef] Herrera, F.; Martinez, L. A 2-tuple fuzzy linguistic representation model for computing with words. IEEE Trans. Fuzzy Syst. 2000, 8, 746–752. Herrera, F.; Martinez, L. An approach for combining linguistic and numerical information based on the 2-tuple fuzzy linguistic representation model in decision-making. Int. J. Uncertain. Fuzz. Knowl.-Based Syst. 2000, 8, 539–562. [CrossRef] Merigó, J.M.; Casanovas, M.; Martínez, L. Linguistic aggregation operators for linguistic decision making based on the Dempster-Shafer theory of Evidence. Int. J. Uncertain. Fuzz. Knowl.-Based Syst. 2010, 18, 287–304. [CrossRef] Wei, G.W. A method for multiple attribute group decision making based on the ET-WG and ET-OWG operators with 2-tuple linguistic information. Expert Syst. Appl. 2010, 37, 7895–7900. [CrossRef] Wei, G.W. Extension of TOPSIS method for 2-tuple linguistic multiple attribute group decision making with incomplete weight information. Knowl. Inf. Syst. 2010, 25, 623–634. [CrossRef] Merigo, J.M.; Casanovas, M. Decision Making with Distance Measures and Linguistic Aggregation Operators. Int. J. Fuzzy Syst. 2010, 12, 190–198. Merigo, J.M.; Gil-Lafuente, A.M.; Zhou, L.G.; Chen, H.Y. Generalization of the linguistic aggregation operator and its application in decision making. J. Syst. Eng. Electron. 2011, 22, 593–603. [CrossRef] Wei, G.W. Grey relational analysis method for 2-tuple linguistic multiple attribute group decision making with incomplete weight information. Expert Syst. Appl. 2011, 38, 4824–4828. [CrossRef] Wei, G.W. Some harmonic averaging operators with 2-tuple linguistic assessment information and their application to multiple attribute group decision making. Int. J. Uncertain. Fuzz. Knowl.-Based Syst. 2011, 19, 977–998. [CrossRef] Wei, G.W.; Zhao, X.F. Some dependent aggregation operators with 2-tuple linguistic information and their application to multiple attribute group decision making. Expert Syst. Appl. 2012, 39, 5881–5886. [CrossRef] Xu, Y.J.; Wang, H.M. Approaches based on 2-tuple linguistic power aggregation operators for multiple attribute group decision making under linguistic environment. Appl. Soft Comput. 2011, 11, 3988–3997. [CrossRef]
Symmetry 2018, 10, 131
38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52.
53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64.
35 of 36
Liao, X.W.; Li, Y.; Lu, B. A model for selecting an ERP system based on linguistic information processing. Inf. Syst. 2007, 32, 1005–1017. [CrossRef] Wang, W.P. Evaluating new product development performance by fuzzy linguistic computing. Expert Syst. Appl. 2009, 36, 9759–9766. [CrossRef] Tai, W.S.; Chen, C.T. A new evaluation model for intellectual capital based on computing with linguistic variable. Expert Syst. Appl. 2009, 36, 3483–3488. [CrossRef] Zhao, X.F.; Li, Q.; Wei, G.W. Some prioritized aggregating operators with linguistic information and their application to multiple attribute group decision making. J. Intell. Fuzzy Syst. 2014, 26, 1619–1630. Jiang, X.P.; Wei, G.W. Some Bonferroni mean operators with 2-tuple linguistic information and their application to multiple attribute decision making. J. Intell. Fuzzy Syst. 2014, 27, 2153–2162. Beliakov, G.; James, S.; Mordelová, J.; Rückschlossová, T.; Yager, R.R. Generalized Bonferroni mean operators in multicriteria aggregation. Fuzzy Sets Syst. 2010, 161, 2227–2242. [CrossRef] Bonferroni, C. Sulle medie multiple di potenze. Boll. Unione Mat. Ital. 1950, 5, 267–270. Xia, M.M.; Xu, Z.; Zhu, B. Generalized intuitionistic fuzzy Bonferroni means. Int. J. Intell. Syst. 2012, 27, 23–47. [CrossRef] Xu, Z.S.; Yager, R.R. Intuitionistic fuzzy Bonferroni means. IEEE Trans. Syst. Man Cybern. 2011, 41, 568–578. Yager, R.R. On generalized Bonferroni mean operators for multi-criteria aggregation. Int. J. Approx. Reason. 2009, 50, 1279–1286. [CrossRef] Wei, G.W. Picture 2-tuple linguistic Bonferroni mean operators and their application to multiple attribute decision making. J. Intell. Fuzzy Syst. 2017, 19, 997–1010. [CrossRef] Tang, X.Y.; Wei, G.W. Models for green supplier selection in green supply chain management with Pythagorean 2-tuple linguistic information. IEEE Access 2018, 6, 18042–18060. [CrossRef] Wei, G.W.; Zhao, X.F.; Lin, R.; Wang, H.J. Uncertain linguistic Bonferroni mean operators and their application to multiple attribute decision making. Appl. Math. Model. 2013, 37, 5277–5285. [CrossRef] Zhu, B.; Xu, Z.S.; Xia, M.M. Hesitant fuzzy geometric Bonferroni means. Inform. Sci. 2012, 205, 72–85. [CrossRef] Zhang, R.; Wang, J.; Zhu, X.; Xia, M.; Yu, M. Some Generalized Pythagorean Fuzzy Bonferroni Mean Aggregation Operators with Their Application to Multiattribute Group Decision-Making. Complexity 2017, 2017, 5937376. [CrossRef] Fang, Z.; Ye, J. Multiple Attribute Group Decision-Making Method Based on Linguistic Neutrosophic Numbers. Symmetry 2017, 9, 111. [CrossRef] Shi, L.; Ye, J. Cosine Measures of Linguistic Neutrosophic Numbers and Their Application in Multiple Attribute Group Decision-Making. Information 2017, 8, 117. Chen, T. The inclusion-based TOPSIS method with interval-valued intuitionistic fuzzy sets for multiple criteria group decision making. Appl. Soft Comput. 2015, 26, 57–73. [CrossRef] Wei, G.; Alsaadi, F.E.; Hayat, T.; Alsaedi, A. Projection models for multiple attribute decision making with picture fuzzy information. Int. J. Mach. Learn. Cybern. 2018, 9, 713–719. [CrossRef] Merigo, J.M.; Gil-Lafuente, A.M. Fuzzy induced generalized aggregation operators and its application in multi-person decision making. Expert Syst. Appl. 2011, 38, 9761–9772. [CrossRef] Wei, G.W.; Gao, H. The generalized Dice similarity measures for picture fuzzy sets and their applications. Informatica 2018, 29, 1–18. [CrossRef] Gao, H.; Wei, G.W.; Huang, Y.H. Dual hesitant bipolar fuzzy Hamacher prioritized aggregation operators in multiple attribute decision making. IEEE Access 2018, 6, 11508–11522. [CrossRef] Merigó, J.M.; Gil-Lafuente, A.M. Induced 2-tuple linguistic generalized aggregation operators and their application in decision-making. Inform. Sci. 2013, 236, 1–16. [CrossRef] Yu, D.J.; Wu, Y.Y.; Lu, T. Interval-valued intuitionistic fuzzy prioritized operators and their application in group decision making. Knowl.-Based Syst. 2012, 30, 57–66. [CrossRef] Gao, H.; Lu, M.; Wei, G.W.; Wei, Y. Some novel Pythagorean fuzzy interaction aggregation operators in multiple attribute decision making. Fundam. Inform. 2018, 159, 385–428. [CrossRef] Wei, G.W.; Alsaadi, F.E.; Hayat, T.; Alsaedi, A. Bipolar fuzzy Hamacher aggregation operators in multiple attribute decision making. J. Intell. Fuzzy Syst. 2018, 20, 1–12. [CrossRef] Wei, G.W. Some similarity measures for picture fuzzy sets and their applications. Iran. J. Fuzzy Syst. 2018, 15, 77–89.
Symmetry 2018, 10, 131
65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77.
36 of 36
Wei, G.W.; Wei, Y. Similarity measures of Pythagorean fuzzy sets based on cosine function and their applications. Int. J. Intell. Syst. 2018, 33, 634–652. [CrossRef] Wei, G.W.; Lu, M. Pythagorean fuzzy power aggregation operators in multiple attribute decision making. Int. J. Intell. Syst. 2018, 33, 169–186. [CrossRef] Ma, Z.M.; Xu, Z.S. Symmetric Pythagorean Fuzzy Weighted Geometric_Averaging Operators and Their Application in Multicriteria Decision-Making Problems. Int. J. Intell. Syst. 2016, 31, 1198–1219. [CrossRef] Wei, G.W.; Lu, M. Pythagorean Fuzzy Maclaurin Symmetric Mean Operators in multiple attribute decision making. Int. J. Intell. Syst. 2018, 33, 1043–1070. [CrossRef] Wei, G.W.; Alsaadi, F.E.; Hayat, T.; Alsaedi, A. Picture 2-tuple linguistic aggregation operators in multiple attribute decision making. Soft Comput. 2018, 22, 989–1002. [CrossRef] Wei, G.W. Picture fuzzy Hamacher aggregation operators and their application to multiple attribute decision making. Fundam. Inform. 2018, 157, 271–320. [CrossRef] Merigo, J.M.; Casanovas, M. Induced aggregation operators in decision making with the Dempster-Shafer belief structure. Int. J. Intell. Syst. 2009, 24, 934–954. [CrossRef] Wei, G.W.; Lu, M.; Tang, X.Y.; Wei, Y. Pythagorean Hesitant Fuzzy Hamacher Aggregation Operators and Their Application to Multiple Attribute Decision Making. Int. J. Intell. Syst. 2018, 1–37. [CrossRef] Wei, G.W.; Gao, H.; Wei, Y. Some q-Rung Orthopair Fuzzy Heronian Mean Operators in Multiple Attribute Decision Making. Int. J. Intell. Syst. 2018. [CrossRef] Xu, Z.; Gou, X. An overview of interval-valued intuitionistic fuzzy information aggregations and applications. Granul. Comput. 2017, 2, 13–39. [CrossRef] Meng, S.; Liu, N.; He, Y. GIFIHIA operator and its application to the selection of cold chain logistics enterprises. Granul. Comput. 2017, 2, 187–197. [CrossRef] Wang, C.; Fu, X.; Meng, S.; He, Y. Multi-attribute decision making based on the SPIFGIA operators. Granul. Comput. 2017, 2, 321–331. [CrossRef] Xu, Z.; Wang, H. Managing multi-granularity linguistic information in qualitative group decision making: An overview. Granul. Comput. 2016, 1, 21–35. [CrossRef] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).