Professional Documents
Culture Documents
DOI:10.3233/JIFS-172009
IOS Press
f
roo
3 Shahzaib Ashrafa,∗ , Saleem Abdullaha , Tahir Mahmoodb , Fazal Ghania and Tariq Mahmoodc
4
a Department of Mathematics, Abdul Wali Khan University, Mardan, Pakistan
5
b Department of Mathematics and Statistics, International Islamic University, Islamabad, Pakistan
c Department of Electronics Engineering, University of Engineering and Technology,
rP
6
Taxila sub campus Chakwal, Pakistan
7
tho
8 Abstract. The key objective of the present proposed work in this paper is introduced a new version of picture fuzzy set so
9 called spherical fuzzy sets (SFS). spherical fuzzy set is a new extension of picture fuzzy sets and Pythagorean fuzzy sets. In
10 spherical fuzzy sets, membership degrees are gratifying the condition 0 P 2 (x) + I 2 (x) + N 2 (x) 1 instead of 0 P(x) +
Au
11 I(x) + N(x) 1 as is in picture fuzzy sets. In this paper, we investigate the basic operations of spherical fuzzy sets and discuss
12 some related results. We extend operational laws to aggregation operators and introduce weighted averaging and weighted
13 geometric aggregation operators based on spherical fuzzy number’s. Further a multi attribute decision making method is
14 developed and these aggregation operators are utilized. Finally, we constructed a numerical approach for implementation of
15 proposed technique.
d
16 Keywords: Spherical fuzzy sets, aggregation operators, multi-attribute decision making problem
cte
36
22 track for treating the multi-attribute decision mak- P(x) and non-membership degree N(x)) to define the 37
23 ing problems. Based on fuzziness circumstances IFS with the condition that 0 P(x) + N(x) 1. 38
24 fuzzy sets (FSs), developed by Zadeh [31], was ini- From last few decades, the IFS has been explored 39
tially used. In FSs each element x of the domain
Un
25
by many researchers and successfully applied to 40
26 set contains only one index namely as degree of many practical fields like medical diagnosis, cluster- 41
27 membership P(x) which oscillate from 0 to 1. Non- ing analysis, decision making and pattern recognition 42
28 membership degree for the FS is straightforward [1, 2, 5, 15, 16, 32, 34]. Later, the degree of 43
29 equivalent to 1 − P(x). However, sometime FS has membership and non-membership in IFSs may be 44
30 some drawbacks for example, it has no ability to show denoted as interval values alternatively by crisp num- 45
∗ Corresponding
bers. So, the interval valued intuitionistic fuzzy sets 46
author. Shahzaib Ashraf, Department of Math-
ematics, Abdul Wali Khan University, Mardan, Pakistan. E-mail:
(IVIFSs), was developed by Atanassov and Gar- 47
1064-1246/18/$35.00 © 2018 – IOS Press and the authors. All rights reserved
2 S. Ashraf et al. / Spherical fuzzy sets and their applications
49 The IFS and IVIFS have been explored by many idea of spherical fuzzy sets (SFSs) is introduced as a
50 researchers like as [4, 5], and widely applied to generalization of PFS. In SFS, membership degrees
51 many areas, such as group decision making [22, 27], are gratifying the condition.
52 similarity measures [11], multi-criteria decision mak-
53 ing [24]. In 2017, Lui [19] introduced the Heronian 0 < P 2 (x) + I 2 (x) + N 2 (x) < 1.
54 aggregation operators of intuitionistic fuzzy num- The objectives of this paper are: (1) to introduce the 92
55 bers, also defined interaction partitioned Bonferroni spherical fuzzy set (2) to define the spherical fuzzy 93
56 mean operators [18] and Interval-valued intuition- numbers (SFNs) and related basic operational iden-
f
94
57 istic fuzzy power Bonferroni aggregation operators tities, (3) to suggest score, accuracy and certainty
roo
95
58 [17] for intuitionistic fuzzy numbers. In 2018, Lui functions for comparison, (4) to propose spherical 96
59 [20] introduced the Partitioned Heronian means aggregation operators and some debate on their prop- 97
60 based on linguistic intuitionistic fuzzy numbers for erties, (5) to demonstrate a MADM method based 98
61 dealing with multi-attribute group decision making on these aggregation operators under spherical fuzzy 99
62 problems. information.
rP
100
63 Since in some real life decision theory the deci- The superfluity of this paper is planned as follows. 101
64 sion makers deal with the situation of particular In Section 2, the basic notion of Intuitionistic fuzzy 102
65 attributes where values of their summation of mem- sets, PFSs and their properties are presented. In Sec- 103
66 bership degrees exceeds 1. In such condition, IFSs tion 3, we introduce the concept of SFSs and their 104
tho
67 has no ability to obtain any satisfactory result. To operational properties. Section 4, consist of the study 105
68 overcome this situation Yager [28] developed the of aggregation operators to aggregate the spherical 106
69 idea of Pythagorean fuzzy set (PyFS) as a gen- fuzzy information. In Sections 5, we proposed the 107
70 eralization of IFS, which satisfies that the value MADM method to deal with spherical fuzzy informa- 108
71 of square summation of its membership degrees is tion and a descriptive example is illustrated to express 109
Au
72 less then or equals to 1. Now the situation where the effectiveness and reliability of the suggested tech- 110
73 the neutral membership degree calculate indepen- nique. The last Section 6, consist of the conclusion 111
74 dently in real life problems, the IFS and PyFS fail of the paper. 112
75 to attain any satisfactory result. Based on these cir-
76 cumstances, to overcome this situation, Cuong and
Kreinovich [9] initiated the idea of picture fuzzy
d
77
2. Preliminaries 113
78 set (PFS). He utilized three index (membership
degree P (x), neutral-membership degree I (x), and
cte
79
The paper gives brief discussion on basic ideas
non-membership degree N (x)) in PFS with the condi-
114
80
associated to IFS and PFS with their operations and
tion that is 0 P (x) + I (x) + N (x) 1. Obviously
115
81
operators. We also discuss, more familiarized ideas 116
82 PFSs is more suitable than IFS and PyFS to deal with
which is utilized in following analysis 117
83 fuzziness and vagueness. H. Garg [12] introduced
rre
88 been explored by many researchers and successfully A = {r, PA (r), NA (r)|r ∈ R} ,
89 applied to many practical fields like strategic deci-
90 sion making, Attribute decision making and pattern is said to be intuitionistic fuzzy set of R, where PA : 119
91 recognition [9, 10, 12, 26]. R → [0, 1] and NA : R → [0, 1] are said to be 120
Un
Sometimes in real life, we face many problems degree of positive-membership of r in R and negative- 121
which cannot be handled by using PFS for example membership degree of r in R espectively. Also PA 122
when P (x) + I (x) + N (x) > 1. In such condition, and NA satisfy the condition: (∀r ∈ R) (0 PA (r) + 123
PFS has no ability to obtain any satisfactory result. To NA (r) 1). 124
is said to be picture fuzzy set of R, where PA : I(x) + N(x) = 1.5 > 1. But by squaring P(x), I(x) 166
R → [0, 1], IA : R → [0, 1] and NA : R → and N(x) their sum becomes less than or equal to one 167
[0, 1] are said to be degree of positive-membership i.e. 0 (0.7)2 + (0.3)2 + (0.5)2 1. This example 168
of r in R, neutral-membership degree of r in R and indicate that SFS is more influential tool than PFS. 169
negative-membership degree of r in R respectively. Also observe that PyFS is also failed to deal such 170
Also PA , IA and NA satisfy the condition: situation because PyFS does not deal with neutral 171
(∀r ∈ R)(0 PA (r) + IA (r) + NA (r) 1). failed to deal such type of information, so this situa- 173
f
tion leads us to propose novel structure of SFSs. This 174
roo
126 Then for 1 − (PA (r) + IA (r) + NA (r)) is said to be example give strength to the novelty and effectiveness 175
127 degree of refusal-membership of r in R. of proposed SFSs. 176
rP
129 In this section, the idea of spherical fuzzy set (SFS),
130 their operations and operators are introduced. The
131 structure of SFSs give shields to the elements, which
tho
132 satisfies or dissatisfies the condition that values must
133 oscillate from 0 to 1. We see in the figure : 1,
134 which specifies the points where structure of Spher-
135 ical fuzzy set give shield to the elements. Also,
136 we familiarized with the contour of this graph in
figure : 3.
Au
137
142
144 between PyFSs and SFSs is that in SFSs we study Now we see the contour of this graph. 177
145 the neutral degree, where as in PyFSs it doesn’t. In
146 PFSs the connection of positive, neutral and negative
147 grades of an object is given in unit close interval but
148 in some cases the sum of the positive, neutral and
rre
165 0.3 and N(x) = 0.5. Then in this case 0 P(x) + contour, we have. 179
4 S. Ashraf et al. / Spherical fuzzy sets and their applications
195
198
f
Suppose that ej = Pej , Iej , Nej and ek =
201
roo
Pek , Iek , Nek are any two SFNs and τ 0. Then 202
(1) τej = 1 − (1 − Pe2j )τ , (Iej )τ , (Nej )τ ; 204
rP
205
tho
180
209
Let the universe set be R. Then the set
Ne2j + Ne2k − Ne2j · Ne2k ; 210
218
220
222
181 are said to SFN and each SFN can be denoted by (3) (ej + ek ) + el = ej + (ek + el ); 223
182 e = Pe , Ie , Ne , where Pe , Ie and Ne ∈ [0, 1], with 224
183 condition that 0 Pe2 + Ie2 + Ne2 1. (4) (ej × ek ) × el = ej × (ek × el ); 225
co
226
186 Pek , Iek , Nek be any two SFNs. Then union, inter- (6) τj ej + τk ej = (τj + τk )ej , τj 0 and τk 0; 229
Un
190 234
f
= 1 − (1 − Pe2j )τj , (Iej )τj , (Nej )τj
roo
Hence, we prove this.
+ 1 − (1 − Pe2j )τk , (Iej )τk , (Nej )τk
237
rP
= Pej , Iej , Nej × Pek , Iek , Nek
= Pej · Pek , Iej · Iek , Ne2j + Ne2k − Ne2j · Ne2k (Iej )τ · (Iek )τ , (Nej )τ · (Nek )τ
τj τk
τ +τ
= Pek · Pej , Iek · Iej , Ne2k + Ne2j − Ne2k · Ne2j = 1 − 1 − Pe2j 1 − Pe2j , Iej j k ,
τ +τ
tho
= ek × ej = R.H.S Nej j k
Hence, we prove this. τj +τk
238
τ +τ
239 The proofs of (3) and (4) are straightforward as (1) = 1 − 1 − Pej 2 , Iej j k ,
and (2). τj +τk
240
Au
241 (5) We have to show that τej + τek = τ(ej + ek ), Nej
242 τ 0. Consider
Furthermore 247
L.H.S = τej + τek
d
τ Nej
Nej · Nek
τ
= 1 − 1 − Pe2j + Pe2k − Pe2j · Pe2k , Hence, we prove this.
τ τ τ 248
L.H.S = ej × ek , τ 0
Un
R.H.S = τ(ej + ek ) = Pej · Pek , Iej · Iek ,
τ
= τ · Pe2j +Pe2k −Pe2j ·Pe2k , Iej ·Iek , Nej ·Nek Ne2j + Ne2k − Ne2j · Ne2k
τ τ
τ = Pej · Pek , Iej · Iek ,
= 1 − 1 − Pe2j + Pe2k − Pe2j · Pe2k ,
τ
τ τ 1 − 1 − Ne2j + Ne2k − Ne2j · Ne2k
Iej · Iek , Nej · Nek
6 S. Ashraf et al. / Spherical fuzzy sets and their applications
× (Pek )τ , (Iek )τ , 1 − (1 − Ne2k )τ play an important role for the ranking of SFNs are; 253
= (Pej )τ · (Pek )τ , (Iej )τ · (Iek )τ ,
3.2.1. Definition
254
f
1 − (1 − N 2 )τ + 1 − (1 − N 2 )τ Let ek = Pek, Iek , Nek be any
e ek SFNs. Then 255
roo
j
Pek +1−Iek +1−Nek
sc(ek ) = = 13 (2 + Pek −
− 1 − (1 − Ne2j )τ 1 − (1 − Ne2k )τ (1) 3 256
τ racy function.
τ
259
rP
1 − 1 − Nej
2 1 − Nek
2 260
= Pej · Pek , Iej · Iek , technique which using for equating the SFNs can be 262
τ described as 263
tho
3.2.2. Definition 264
0 and τk 0. Consider using the Definition [3.2.1.], equating technique can 267
= (Pej )τj , (Iej )τj , 1 − (1 − Ne2j )τj cr(ej ) > cr(ek ), then ej > ek. 272
(d) If sc(ej ) = sc(ek ), ac(ej ) = ac(ek ) and
d
273
× (Pej )τk , (Iej )τk , 1 − (1 − Ne2j )τk cr(ej ) = cr(ek ), then ej = ek. 274
= (Pej )τj · (Pej )τk , (Iej )τj · (Iej )τk ,
cte
1 − (1 − N 2 )τj + 1 − (1 − N 2 )τk 4. Spherical aggregation operators
ej ej 275
− 1 − (1 − Ne2j )τj 1 − (1 − Ne2j )τk This section gives some discussion about weighted 276
rre
τj τk
1 − 1 − Nej2 1 − Nej
2 4.1. Spherical fuzzy number weighted averaging 279
co
285
τ = {τ1 , τ2 , .. . , τn } be the weight vector
In which (c) Now we have to prove that outcome is true for
of e = Pek , Iek , Nek , k ∈ N, with τk 0 and n = z + 1, by utilizing the (a) & (b) we have
286
n k
287
k=1 τk = 1.
z
SFNWAA (e1 , e2 , . . . , ez , ez+1 ) = τk ek + τz+1 ez+1
288
4.1.2. Theorem k=1 τk
289 Let ek = Pek , Iek , Nek , (k ∈ N) be any collec- = 1 − zk=1 (1 − Pe2k )τk , zk=1 Iek , zk=1 (Nek )τk +
290 tion of SFNs. Then by utilizing the Definition [4.1.1.] τz+1 τz+1
291 and operational properties of SFNs, we can obtain the 1 − (1 − Pe2z+1 )τz+1 , Iez+1 , Nez+1
f
292 following outcome.
1 − zk=1 (1 − Pe2k )τk + 1 − (1 − Pe2z+1 )τz+1
roo
,
293 SFNWAA (e1 , e2 , . . . , en ) = ( 1 − nk=1 (1 − Pe2k )τk ), = − 1 − zk=1 (1 − Pe2k )τk 1 − (1 − Pe2z+1 )τz+1
τk τz+1 τz+1
zk=1 Iek · Iez+1 , zk=1 (Nek )τk · Nez+1
294 nk=1 (Iek )τk , nk=1 (Nek )τk ,
τk
1 − zk=1 (1 − Pe2k )τk (1 − Pe2z+1 )τz+1 , zk=1 Iek
τz+1
rP
295 = τz+1 , zk=1 (Nek )τk· Nez+1
296 Where τ = {τ1 , τ2 , . .. , τn } be the weight vector · Iez+1
τk
297
nek = Pek , Iek , Nek , k ∈ N, with τk 0 and
of
= 1 − zk +1 2 τk z +1
=1 (1 − Pek ) , k =1 Iek , kz +1
=1 (Nek )
τk
298
k=1 τk = 1.
304
tho
299 Proof. We done the prove by utilizing the technique
300 of mathematical induction. Therefore, we follow as Outcome is satisfying for n = z + 1. Thus, outcome 305
(a) For n = 2, since is satisfied for whole n. Therefore 306
τ τ 307
τ1 e1 = (1 − (1 − Pe21 )τ1 , Ie1 1 , Ne1 1 )
Au
SFNWAA (e1 , e2 , . . . , en ) 308
and = 1 − nk=1 (1 − Pe2k )τk , nk=1 (Iek )τk , nk=1 (Nek )τk 309
τ τ
τ2 e2 = (1 − (1 − Pe22 )τ2 , Ie2 2 , Ne2 2 ) 310
Then
d
SFNWAA (e1 , e2 ) = τ1 e1 + τ2 e2
cte
= (1 − (1 − Pe21 )τ1 , Ie1 1 , Ne1 1 ) There are some properties which are fulfilled by 313
τ τ the SFNWAA operator obviously.
+ (1 − (1 − Pe22 )τ2 , Ie2 2 , Ne2 2 ) 314
Pe Ie Ne
maxk k , mink k , mink k for whole k ∈ N, there-
2k=1 (Nek )τk
fore e− +
k ⊆ SFNOWAA (e1 , e2 , . . . , en ) = ek .
(b) Suppose that outcome is true for n = z that is, e− +
k ⊆ SFNWAA (e1 , e2 , . . . , en ) = ek .
301
302 SFNWAA (e1 , e2 , . . . , ez ) = (c) Monotonically: Let ȩk = Pȩk , Iȩk , Nȩk ,
τ (k ∈ N) be any collection of SFNs. If it satisfies that
1 − zk=1 (1 − Pe2k )τk , zk=1 Iek k , ek ⊆k for whole k ∈ N, then
,
303
315 337
316 4.1.4. Definition
Let ek = Pek , Iek , Nek (kεN) be a collection of The spherical fuzzy weighted averaging operator 338
SFNs, and SFNOWAA : SFN n → SFN, then the only considers importance of the aggregated spheri- 339
SFNOWAA describe as, cal fuzzy sets themselves. The spherical fuzzy OWA 340
SFNOWAAw (e1 , e2 , . . . , en ) = nk=1 τk eη(k) the ranking order of the aggregated spherical fuzzy 342
f
318
we may define the following spherical fuzzy hybrid 345
. . . eη(n) . τ = {τ1 , τ2 , . . . , τn } is the weight vec-
roo
319
weighted averaging operator.
tors such that τk 0 (k ∈ N) and nk=1 τk = 1.
346
320
4.1.7. Definition
347
321
4.1.5. Theorem Let ek = Pek , Iek , Nek (kεN) be a collection of
322 Let ek = Pek , Iek , Nek (k ∈ N) be a collection SFNs, and SFNHOWAA : SFN n → SFN, then
rP
323 of SFNs. Then by utilizing the definition [4.1.4.] and the SFNHOWAA describe as,
324 operational properties of SFNs, we can obtain the
325 following outcome. SFNHOWAAw (e1 , e2 , . . . en ) = nk=1 τk eη(k) ,
SFNOWAAw (e1 , e2 , . . . , en )
326
τk with dimensions n, where kth biggest weighted
348
tho
1 − nk=1 (1 − Pe2η(k) )τk , nj=1 Ieη(k) , value is eη(k) and ek ek = nτk ek , k εN , 349
327 = τk τ = {τ1 , τ2 , . . . , τn } is the weight vectors
nj=1
350
Neη(k)
such that τk 0 (kεN) and nk=1 τk = 1. Also 351
328
ω = (ω1 , ω2 , . . . , ωn ) is the associated weight 352
329 where eη(k) is kth largest value consequently by total vector such that ωk 0 (k ∈ N) and nk=1 ωk = 1,
Au 353
330 order eη(1) eη(2) . . . eη(n) . and balancing coefficient is n. 354
333 procedure is eliminated here. Let ek = Pek , Iek , Nek (k ∈ N) be a collection 356
357
following outcome.
cte
is nj=1 Ie , nj=1 Ne
η(k) η(k)
SFNOWAA (e1 , e2 , . . . , en ) = e.
where eη(k) is kth biggest value consequently by 362
co
(b) Boundedness: Let ek = Pek , Iek , Nek , total order eη(1) eη(2) .. . eη(n) and kth biggest
363
(k ∈ N) be any collection of SFNs. Assuming weighted value is eη(k) , ek ek = nτk ek , k εN . 364
− Pek Iek Nek
that ek = mink , mink , maxk and e+
k =
Pek Iek Nek
Proof. Theorem [4.1.8.] takes the form by utilizing 365
Un
maxk , mink , mink for whole k ∈ N, there- the technique of mathematical induction on n and 366
e− +
k ⊆ SFNOWAA (e1 , e2 , . . . , en ) = ek . 4.1.9. Properties
368
(c) Monotonically: Let ȩk = Pȩk , Iȩk , Nȩk , There are some properties which are fulfilled by 369
(k ∈ N) be any collection of SFNs. If it satisfies that the SFNHOWAA operator obviously. 370
SFNWAA (e1 , e2 , . . . , en ) = SFNOWAA ȩ1 , ȩ2 , . . . , ȩn . (k ∈ N) be any collection of SFNs. If all of 372
S. Ashraf et al. / Spherical fuzzy sets and their applications 9
373 ek = Pek , Iek , Nek , (k ∈ N) are identical. following outcome. 399
n (P )τk , n (I )τk ,
SFNHOWAA (e1 , e2 , . . . , en ) = e. k=1 ek k=1 ek
= ,
1 − nk =1 (1 − Ne2k )τk
401
375 (b) Boundedness: Let ek = Pek , Iek , Nek ,
(k ∈ N) be any
f
376
collection of SFNs. Assuming
Pek Iek Ne Where τ = {τ1 , τ2 , . .. , τn } be the weight vector
roo
−
that ek = mink , mink , maxk k
402
377 and
P I N
nek = Pek , Iek , Nek , k ∈ N, with τk 0 and
of 403
e+
e e e
k = maxk , mink , mink k=1 τk = 1.
k k k
378 for whole 404
379 k ∈ N, therefore
Proof. We done the prove by utilizing the technique 405
rP
406
τ1 τ1
(c) Monotonically: Let ȩk = Pȩk , Iȩk , eτ11 = Pe1 , Ie1 , (1 − (1 − Ne21 )τ1 )
380
tho
381 Nȩk , (k ∈ N) be any collection of SFNs. If
382 it satisfies that ek ⊆ ȩk for whole k ∈ N, then and
τ
383
n
τ τ
τ = 2k=1 Pek k , 2k=1 Iek k ,
SFNWGA (e1 , e2 , . . . , en ) = ekk ,
k=1 1 − 2k =1 (1 − Ne2k )τk
Un
392
τ = {τ1 , τ2 , .. . , τn } be the weight vector
In which
393 of
n ek = Pek , Iek , Nek , k ∈ N, with τk 0 and
(b) Suppose that outcome is true for n = z that is,
k=1 τk = 1.
409
394
(c) Now we have to prove that outcome is true for 4.2.4. Definition
428
n = z + 1, by utilizing the (a) & (b) we have Let ek = Pek , Iek , Nek , (k ∈ N) be any collec-
z tion of SFNs and SFNOWGA : SFN n → SFN,
τ τz+1
SFNWGA (e1 , e2 , . . . , ez , ez+1 ) = ekk + ez+1 then SFNOWGA operator describe as,
k=1
τk τk
= zk=1 Pek , zk=1 Iek , 1 − zk=1 (1 − Ne2k )τk + SFNOWGAw (e1 , e2 , . . . , en ) = nk=1 eτη(k)
k ,
τ 1 τ1
Pe 1 , Ie 1 , (1 − (1 − Ne21 )τ1 )
τk τz+1 τk τz+1 with dimensions n, where kth biggest weighted 429
zk=1 Pek · Pez+1 , zk=1 Iek · Iez+1 , value is eη(k) consequently by total order eη(1)
f
430
roo
z
1 − k=1 (1 − Nek ) + 1 − (1 − Nez+1 )
τ τ
2 2 431
k z+1
=
− 1 − zk=1 (1 − Ne2k )τk
1 − (1 − Ne2z+1 )τz+1
weight vector of ek = Pek , Iek , Nek , k ∈ N, with 432
τk τz+1 τk τz+1
zk=1 Pek · Pez+1 , zk=1 Iek · Iez+1 ,
= 4.2.5. Theorem
434
1 − zk=1 (1 − Ne2k )τk (1 − Ne2z+1 )τz+1 Let ek = Pek , Iek , Nek , (k ∈ N) be any collec-
rP
435
τk τk tion of SFNs. Then by utilizing the definition [4.2.4.] 436
= z+1 , z+1 , 1 − z+1 2 τk
k=1 Pe k k=1 Ie k k=1 (1 − Nek ) and operational properties of SFNs, we can obtain the 437
tho
414
SFNOWGAw (e1 , e2 , . . . , en ) 439
come is satisfied for whole n. Therefore
n P τk , n (I )τk ,
415
419 which done the proof. Proof. Theorem [4.2.5.] takes the form by utilizing 443
(a) Idempotency: Let ek = Pek , Iek , Nek , There are some properties which are fulfilled by 447
(k ∈ N)
be any collection
of SFNs. If all of 450
rre
SFNOWGA (e1 , e2 , . . . , en ) = e.
Pek Iek Nek
maxk , mink , mink for whole k ∈ N, there-
fore (b) Boundedness: Let ek = Pek , Iek , Nek , 453
Un
(c) Monotonically: Let ȩk = Pȩk , Iȩk , Nȩk , Pek Iek Nek
423
+
ek = maxk , mink , mink for whole
(k ∈ N) be any collection of SFNs. If it satisfies that
456
424
(c) Monotonically: Let ȩk = Pȩk , Iȩk ,
458
f
fore
roo
464 4.2.7. Definition
e− +
k ⊆ SFNHOWGA (e1 , e2 , . . . , en ) = ek .
Let ek = Pek , Iek , Nek , (k ∈ N) be any collec-
tion of SFNs and SFNHOWGA : SFN n → SFN, (c) Monotonically: Let ȩk = Pȩk , Iȩk , Nȩk , 492
then SFNOWGA operator describe as, (k ∈ N) be any collection of SFNs. If it satisfies that 493
rP
τ
ek ⊆ ȩk for whole k ∈ N, then 494
SFNHOWGAw (e1 , e2 , . . . en ) = nj=1 eη(k)
k
,
SFNHOWGA (e1 , e2 , . . . , en ) 495
465 with dimensions n, where kth biggest weighted
466 value is eη(k) and ek ek = nτk ek , k εN , = SFNHOWGA ȩ1 , ȩ2 , . . . , ȩn . 496
tho
467 τ = {τ1 , τ2 , . . . , τn } be the weight vector of
468 ek = Pek , Iek , Nek , k ∈ N, with τk 0 and
n
469
k=1 τk = 1. Also ω = (ω1 , ω2 , . . . , ωn ) is the
5. MADM method utilizing spherical
associated weight vector such that ωj 0 (jεN) and
497
470
aggregation operators
nj=1 ωj = 1, and balancing coefficient is n.
498
471
Au
This section proposes the technique to solve
472 4.2.8. Theorem
the MADM problems by utilizing the spherical
473 Let ek = Pek , Iek , Nek , (k ∈ N) be any collec- aggregation operators. For a MADM problem,
474 tion of SFNs. Then by utilizing the definition [4.2.7.] assuming that C = {c1 , c2 , . . . , cm } be any finite
475 and operational properties of SFNs, we can obtain the collection of m alternatives, G = {g1 , g2 , . . . , gn }
d
η(k)
481
c2 Pe21 , Ie21 , Ne21 Pe22 , Ie22 , Ne22 ... Pe2n , Ie2n , Ne2n
482 Proof. Theorem [4.2.8.] takes the form by utilizing
. . . . .
483 the technique of mathematical induction on n. and . . . . .
484 procedure is eliminated here. . . . . .
Un
cm Pem1 , Iem1 , Nem1 Pem2 , Iem2 , Nem2 ... Pemn , Iemn , Nemn
485 4.2.9. Properties
486 There are some properties which are fulfilled by 499
487 the SFNHOWGA operator obviously. be the DM, where Pejk , Iejk , Nejk are the 500
488 (a) Idempotency: Let ek = Pek , Iek , Nek , information of every alternative ci (i = 1, 2, . . . , m) 502
489 (k ∈ N)
be any collection
of SFNs. If all of on attribute gi (i = 1, 2, . . . , n) . If τ = 503
490 ek = Pek , Iek , Nek , (k ∈ N) are identical. {τ1 , τ2 , . . . , τn } be the weight vector of attribute, 504
491 Then there is with τk 0, nk=1 τk = 1 and the weight vector of 505
12 S. Ashraf et al. / Spherical fuzzy sets and their applications
506 DMs is Q = q1 , q2 , . . . , qp , with qk 0 and environmental impact. The weight vector of the
p
k=1 qk = 1. attributes is τ = (0.35, 0.25, 0.40), and correspond-
ing associated weight are ω = {0.5, 0.2, 0.3} . Now
507
508 Then, listed below the main technique of handling we can calculate the following spherical fuzzy num-
509 the MADM problems: ber decision matrix as
510 Step 1. In extensively, we have two kinds of cri- g1 g2 g3
511 terion one is said to be positive criteria and other c1 0.6, 0.6, 0.2 0.8, 0.4, 0.1 0.8, 0.2, 0.3
512 one said to be negative criteria. For negative crite- c2 0.8, 0.3, 0.2 0.5, 0.4, 0.3 0.4, 0.3, 0.4
f
513 rion, we need to modify the negative criteria into c3 0.4, 0.8, 0.2 0.6, 0.5, 0.2 0.5, 0.3, 0.6
roo
514 positive criteria. In this step we construct the
spheri-
c4 0.7, 0.1, 0.7 0.7, 0.5, 0.1 0.7, 0.3, 0.5
515
s
cal fuzzy decision making matrices, D = bjk s
m×n 546
516 (s = 1, 2, . . . n) for decision. If the criteria have
517 two types, such as positive (benefit) criteria and Step 1. Since the attributes are same type, so there is 547
518 negative (cost) criteria, then the spherical fuzzy no need to be normalized. 548
rP
519 decision matrices, Ds = bjk s can be converted Step 2. (Case:1) Utilizing the SFNWAA Operator to
m×n
into the normalized find out every value of the alternative ci as,
520
spherical fuzzy decision matri-
s
ces, R R = rjk s s =
(s = 1, 2, . . . n) where rjk
521 e1 = 0.748, 0.349, 0.197
m×n
s e2 = 0.627, 0.322, 0.292
tho
bjk for benefit criteria
s is represents the
and bjk
522
s
bjk for cost criteria e3 = 0.499, 0.480, 0.310
523 compliment of the bjks . If all the criteria have the same e4 = 0.700, 0.232, 0.376
524 type, then there is no need of normalization.
(Case:2) Utilizing the SFNOWAA Operator to find
Au
525 Step 2. We accumulate the SFNs for every decision out every value of the alternative as, Firstly we need
526 maker which are given, then use the SFNWAA and score function of the each alternative
527 SFNWGA operators, discussed in Theorem [4.1.2. & g1 g2 g3
528 4.2.2.] to aggregate the spherical fuzzy information. c1 sc(c1 g1 ) = 0.60 sc(c1 g2 ) = 0.76 sc(c1 g3 ) = 0.76
529 Which help us to choose the best alternate in the set
c2 sc(c2 g1 ) = 0.76 sc(c2 g2 ) = 0.60 sc(c2 g3 ) = 0.56
d
530 of alternatives.
c3 sc(c3g1 ) = 0.46 sc(c3 g2 ) = 0.63 sc(c3 g3 ) = 0.53
Step 3. Calculating the score, accuracy and cer-
cte
542 5.1. A Numerical example out every values of the alternative as,
543 This section describes a MADM problem, which is e1 = 0.800, 0.274, 0.197
544 utilized to illustrate the pertinence and effectiveness e2 = 0.700, 0.232, 0.376
545 for the procedure of decision making problems. e3 = 0.535, 0.434, 0.263
There are four manageable alternatives, (a) c1 is
car company; (b) c2 is food company; (c) c3 is e4 = 0.468, 0.454, 0.343
a computer company; (d) c4 is an arms company.
According to the attributes company takes the deci- (Case:3) Utilizing the SFNHOWAA Operator to
sion, (a) g1 is the risk; (b) g2 is the growth; (c) g3 is the find out every value of the alternative as, Firstly we
S. Ashraf et al. / Spherical fuzzy sets and their applications 13
f
eη(3) 3×0.5 × 0.4, 0.8, 0.2 3×0.2 × 0.6, 0.5, 0.2 3×0.3 × 0.5, 0.3, 0.6 Step 4. (Case:1) Now rank the all the alternative 552
roo
eη(4) 3×0.5 × 0.7, 0.1, 0.7 3×0.2 × 0.7, 0.5, 0.1 3×0.3 × 0.7, 0.3, 0.5 by utilizing the comparison technique in Definition 553
We multiply these scaler as in define multiplication sc(e1 ) = 0.73 > sc(e4 ) = 0.70 > sc(e2 ) 555
rP
eη(1) 0.69, 0.46, 0.08 0.67, 0.57, 0.25 0.77, 0.23, 0.33 Thus, according to the scoring function rank- 557
eη(2) 0.88, 0.16, 0.08 0.39, 0.57, 0.48 0.38, 0.33, 0.43 ing method of spherical fuzzy sets for SFNWAA 558
eη(3) 0.47, 0.71, 0.08 0.48, 0.65, 0.38 0.47, 0.33, 0.63 operator, the ranking order of the suppliers ei (i = 559
eη(4) 0.79, 0.03, 0.58 0.57, 0.65, 0.25 0.67, 0.33, 0.53 1, 2, 3, 4) is generated as follows: e1 > e4 > e2 > 560
tho
e3 . The best supplier for the enterprise is e1 . 561
551 (Case:2) Now rank the all the alternative by utiliz- 562
Now by utilizing the SFNHOWAA Operator to find ing the comparison technique in Definition [3.2.2.] 563
out every value of the alternative as, for SFNOWAA operator as, 564
567
and certainty function values respectively by utilizing (Case:3) Now rank the all the alternative by utiliz- 568
the Definition [3.2.1.] as, ing the comparison technique in Definition [3.2.2.] 569
sc(e2 ) = 0.66 ac(e2 ) = 0.34 cr(e2 ) = 0.63 0.66 > sc(e3 ) = 0.56
sc(e3 ) = 0.57 ac(e3 ) = 0.19 cr(e3 ) = 0.50 we obtained the result which is
sc(e4 ) = 0.70 ac(e4 ) = 0.32 cr(e4 ) = 0.70 e1 > e2 > e4 > e3
co
(Case:3) Now we find out the score, accuracy and e1 from the set of alternatives. As SFSs is generalized 575
certainty function values respectively by utilizing the as compared to FSs, IFSs, PyFS and PFSs and have 576
14 S. Ashraf et al. / Spherical fuzzy sets and their applications
577 the ability to deal with real life problems more effec- Step 2. Utilizing the SFNWAA operator to find out
578 tively than the existing concepts as it is discussed every value of the alternative ci as,
579 to evaluation investment company to invest money
580 problem is effectively applied. e1 = 0.658, 0.349, 0.332
e2 = 0.602, 0.361, 0.331
e3 = 0.512, 0.480, 0.367
581 5.2. Validity and reliability test
e4 = 0.636, 0.232, 0.430
f
582 In particularly, to track the best capable alternative
Step 3. Now we find out the score, accuracy and
roo
583 from given group decision matrices are not probable
certainty function values respectively by utilizing the
584 in practically. Wang and Triantaphyllou [26] initiated
Definition [3.2.1.] as,
585 the test to evaluate the reliability and validity of the
586 MADM techniques. These testing criteria are follow sc(e1 ) = 0.66 ac(e1 ) = 0.32 cr(e1 ) = 0.65
587 as sc(e2 ) = 0.63 ac(e2 ) = 0.27 cr(e2 ) = 0.60
rP
588 Test Criteria No:1. An operative and valid MADM sc(e3 ) = 0.55 ac(e3 ) = 0.14 cr(e3 ) = 0.51
589 technique is that if we are substituting a non-optimal sc(e4 ) = 0.65 ac(e4 ) = 0.20 cr(e4 ) = 0.63
590 alternative in worse alternative by showing without
591 any change then this indication of the best capable Step 4. Now rank the all the alternative by utilizing
tho
609
592 alternative without shifting the comparative status of the comparison technique in Definition [3.2.2.] for 610
600 613
601 rank of un-decomposed problem. approach utilizing Test Criteria No:2 & Test Cri- 614
602 615
603 gation operator constructed by MADM technique is To test validity and reliability of suggested 616
604 test by utilizing these testing criteria. technique by utilizing test criterion 2 and 3, 617
605 Validity and Reliability Test for suggested smaller sub-problems as {e1 , e2 , e3 }, {e2 , e3 , e4 } 619
rre
606 approach utilizing Test Criteria No:1. and {e2 , e4 , e1 }. Then by following stages of the sug- 620
To examine effectiveness of the suggested gested technique, accordingly to each sub-problem 621
approach utilizing test criterion no 1, then these we suggest rank to their correspondences as e1 > 622
decision matrices are attained by exchanging the e4 > e2 , e4 > e2 > e3 and e1 > e2 > e3 respec- 623
and non-membership (worse alternative) in the orig- attain e1 > e4 > e2 > e3 that is similar with original 625
inal decision matrices as follows, after that original MADM technique. Hence it displays the transitive 626
decision matrices are transform property. So, under test criterion 2 and 3 recognized 627
Un
0.6, 0.6, 0.2 0.1, 0.4, 0.8 0.8, 0.2, 0.3 nique is valid and reliable. 629
607 Step 1. Since the attributes are uniform so there is conductive using Pythagorean fuzzy TOPSIS method 633
608 no need to normalize. [30] and aggregation operators of PFS [12]. 634
S. Ashraf et al. / Spherical fuzzy sets and their applications 15
635 A Comparison analysis with the Pythagorean satisfactory results in such situations where sum of 683
636 fuzzy TOPSIS. membership degrees is greater than 1, and chose the 684
637 These two approaches are valid to solve the best alternative from decision making problems. 685
f
687
643 the farthest distance from the negative ideal solution dynamic role throughout the decision making
roo
688
644 (NIS). procedure and hence in this track, the comparative 689
645 According to Zhang and Xu [33], the first step is importance of the criteria remains unreformed in 690
646 to identify the Pythagorean fuzzy positive ideal solu- modified problems, then we proposed SFNWAA and 691
647 tion (PIS) and the Pythagorean fuzzy negative ideal SFNWGA operators has been executed to invention 692
648 solution (NIS) of the decision matrix, which are the best alternative. Almost all the researchers have
rP
693
649 P +
= {0.7665, 0.4530 , 0.8185, 0.4125 , operated the IFS by considering the positive and 694
650 0.7635, 0.6165 , 0.8176, 0.5144} detected that in some circumstances, like in situation 696
651 of voting, human thoughts including more degrees 697
tho
as, yes, neutral, no, refusal, and in situation where
P − = {0.5375, 0.7254 , 0.4976, 0.7996 ,
698
652
the sum of membership degree is greater than 1, 699
653 0.4124, 0.7598 , 0.4884, 0.8439} which cannot be exactly characterized in IFS and 700
660 we utilize positive, neutral and negative member- established for ranking the dissimilar alternatives by 708
661 ship degrees that’s square sum is less or equal to 1. utilizing spherical fuzzy environment. The suggested 709
cte
662 Pythagorean fuzzy TOPSIS method fail to solve such technique has been demonstrated with a descriptive 710
663 problems. So, our proposed technique is reliable and example for viewing their effectiveness as well as 711
664 valid for such types of alternatives. reliability. A test to check the reliability and validity 712
714
666 Sets.
operators provide a new direction to the information 715
667 Picture fuzzy set initialized by B. C. Cuong [9]
measure theory and give a new easier track to grip 716
668 and Picture fuzzy aggregation operators which were
the uncertainties throughout the decision-making 717
669 introduced by Harish Garg [12] are listed as, (1)
procedure. 718
co
676 matrices. fuzzy sets, Fuzzy Sets and Systems 61(2) (1994), 137–142. 721
[2] K. Atanassov, Intuitionistic fuzzy sets, Heidelberg: 722
677 The above proposed aggregation operators for PFS Springer, 1999. 723
678 deals with that real-life problem where sum of mem- [3] K. Atanassov, Intuitionistic fuzzy sets, Fuzzy Sets and Sys- 724
679 bership degrees are equals to 1. In situation where tems 20 (1986), 87–96. 725
[4] K. Atanassov and G. Gargov, Interval valued intuitionistic 726
680 sum of membership degrees exceeds 1, PFS fails
fuzzy sets, Fuzzy Sets and Systems 31 (1989), 343–349. 727
681 to give satisfactory results in such situation. On the [5] K. Atanassov, Remark on intuitionistic fuzzy numbers, 728
682 other hand, technique proposed in this paper can give Notes on Intuitionistic Fuzzy Sets 13 (2007), 29–32. 729
16 S. Ashraf et al. / Spherical fuzzy sets and their applications
730 [6] G. Beliakov, H. Bustince, D.P. Goswami, U.K. Mukherjee with multi-attribute group decision making, Applied Soft 781
731 and N.R. Pal, On averaging operators for Atanassov’s intu- Computing 62 (2018), 395–422. 782
732 itionistic fuzzy sets, Information Sciences 181(6) (2011), [21] A.A. Salama and F. Smarandache, Neutrosophic crisp prob- 783
733 1116–11124. ability theory and decision making process, Critical Review 784
734 [7] S. Broumi and F. Smarandache, Single valued neutrosophic XII (2016), 34–48. 785
735 trapezoid linguistic aggregation operators based multi- [22] C. Tan, A multi-criteria interval-valued intuitionistic 786
736 attribute decision making, Bulletin of Pure and Applied fuzzy group decision making with Choquet integral-based 787
737 Sciences 33E(2) (2014), 135–155. TOPSIS, Expert Systems with Applications 38 (2011), 788
738 [8] S. Broumi, J. Ye and F. Smarandache, An extended top- 3023–3033. 789
739 ics method for multiple attribute decision making based on [23] Z.P. Tian, J. Wang, H.Y. Zhang, X.H. Chen and J.Q. Wang, 790
f
740 interval neutrosophic uncertain linguistic variables, Neutro- Simplied neutrosophic linguistic normalized weighted 791
roo
741 sophic Sets and Systems 8 (2015). Bonferroni mean operator and its application to multi- 792
742 [9] B.C. Cuong, Picture fuzzy sets first results. part 1, Seminar criteria decision-making problems, Filomat 30(12) (2016), 793
743 Neuro-Fuzzy Systems with Applications, Preprint 03/2013, 3339–3360. 794
744 Institute of Mathematics, Hanoi, 2013. [24] H. Wang and Y.Q. Zhang, Multi-criteria decision-making 795
745 [10] B.C. Cuong, Picture fuzzy sets, journal of computer science approach based on Atanassov’s intuitionistic fuzzy sets with 796
746 and cybernetics 30(4) (2001), 409–420. incomplete certain information on weights, IEEE Transac- 797
rP
747 [11] B. Farhadinia, A theoretical development on the entropy tions on Fuzzy Systems 21 (2013), 510–515. 798
748 of interval-valued fuzzy sets based on the intuitionis- [25] E. Wang and E. Triantaphyllou, Ranking irregularities when 799
749 tic distance and its relationship with similarity measure, evaluating alternatives by using some ELECTRE methods, 800
750 Knowledge-Based Systems 39 (2013), 79–84. Omega Int J Manag Sci 36 (2008), 45–63. 801
751 [12] H. Garg, Some Picture fuzzy aggregation operators and [26] G. Wei, Some cosine similarity measures for picture fuzzy 802
752 their applications to multicriteria decision-making, Ara- sets and their applications to strategic decision making, 803
tho
753 bian Journal for Science and Engineering 42(12) (2017), informatica 28(3) (2017), 547–564. 804
754 5275–5290. [27] Z.S. Xu, Approaches to multiple attribute group decision 805
755 [13] C.L. Hwang and K. Yoon, Multiple Attribute Decision Mak- making based on intuitionistic fuzzy power aggregation 806
756 ing Methods and Applications, Springer, New York, NY, operators, Knowledge-Based Systems 24 (2011), 749–760. 807
757 1981. [28] R.R. Yager, Pythagorean fuzzy subsets, In: Proc Joint IFSA 808
758 [14] H.W. Liu and G.J. Wang, Multi-criteria methods based World Congress and NAFIPS Annual Meeting, Edmonton, 809
Au
759 on intuitionistic fuzzy sets, European Journal Operational Canada, 2013, pp. 57–61. 810
760 Research 179(1) (2007), 220–233. [29] R.R. Yager and A.M. Abbasov, Pythagorean membership 811
761 [15] P. Liu and S.M. Chen, Multi-attribute group decision mak- grades, complex numbers, and decision making, Interna- 812
762 ing based on intuitionistic 2-tuple linguistic information, tional Journal of Intelligent System 28 (2014), 436–452. 813
763 Information Sciences 430–431 (2018), 599–619. [30] Y. Yang, H. Ding, Z.S. Chen and Y.L. Li, A note on exten- 814
764 [16] P. Liu and P. Wang, Some q-Rung orthopair fuzzy aggre- sion of TOPSIS to multiple criteria decision making with 815
765 gation operators and their applications to multiple-attribute pythagorean fuzzy sets, International Journal of Intelligent 816
d
766 decision making, International Journal of Intelligent Sys- System (2015), 1–5. 817
767 tems 33(2) (2018), 259–280. [31] L.A. Zadeh, Fuzzy sets, Information and Control 8(3) 818
cte
768 [17] P. Liu and H. Li, Interval-valued intuitionistic fuzzy power (1965), 338–356. 819
769 Bonferroni aggregation operators and their application [32] H. Zhao, Z.S. Xu, M.F. Ni and S.S. Liu, Generalized aggre- 820
770 to group decision making, Cognitive Computation, 9(4) gation operator for intuitionistic fuzzy sets, International 821
771 (2017), 494–512. Journal of Intelligent Systems 25(1) (2010), 1–30. 822
772 [18] P. Liu, S.M. Chen and J. Liu, Multiple attribute group [33] X.L. Zhang and Z.S. Xu, Extension of TOPSIS to multi- 823
773 decision making based on intuitionistic fuzzy interaction ple criteria decision making with Pythagorean fuzzy sets, 824
rre
774 partitioned Bonferroni mean operators, Information Sci- International Journal of Intelligent Systems 29 (2014), 825
775 ences 411 (2017), 98–121. 1061–1078. 826
776 [19] P. Liu and S.M. Chen, Group decision making based [34] L.G. Zhou, Z.F. Tao, H.Y. Chen and J. Liu, Continuous 827
777 on Heronian aggregation operators of intuitionistic fuzzy interval-valued intuitionistic fuzzy aggregation operators 828
778 numbers, IEEE Transactions on Cybernetics 47(9) (2017), and their applications to group decision making, Appl Math 829
co