Game-Theoretic Applications of Pentagonal Fuzzy Numbers through New Representation and Defuzzification Schemes

  • Ganesh KumarDepartment of Mathematics, University of Rajasthan, Jaipur-302004, India.
  • Vinod JangidDepartment of Mathematics, University of Rajasthan, Jaipur-302004, India.
  • Govind Shay SharmaDepartment of Mathematics, Acharya Narendra Dev College University of Delhi, India.

Abstract

in the uncertainty theory. [11, 12] analyzed numerous procedures. Type 2 fuzzy sets

or interval-valued fuzzy sets designed by [29]. [3, 4] submitted intuitionistic fuzzy sets.

[27] evolved fuzzy multi-sets. [24] delivered the concept of neutrosophic fuzzy sets.

[5] specified the probability distribution of the objective function in light of parame-

ters expressed as fuzzy triangular numbers. [8] suggested a new technique for ranking

fuzzy numbers based on length. [14] supplied the theoretical basis of fuzzy numbers,

which was associated with the possibility and Dempster-Shafer theories and conveyed

various types of fuzzy number expressions. [9] suggested a novel arithmetic notion

and process for conducting arithmetic operations on triangular-fuzzy numbers. [16] depicted fuzzy arithmetic operations to eradicate some of the ambushes of the classical

approach. From the perspective of analytical geometry, [25] designated the centroid

formulae for fuzzy numbers. [2] offered a defuzzification using the two fuzzy number’s

distance minimizer. Several modes of ordering trapezoidal fuzzy numbers were con-

ceived by [1]. [19] submitted trapezoidal intuitionistic fuzzy numbers (IF numbers)

and revealed some of its operations. Some further arithmetic operators of the trape-

zoidal fuzzy numbers utilized in [26] suggested an efficient explanation for fuzzy risk

investigation. [7] furnished a novel hazard analysis procedure based on a new tool for

generalized hierarchies. A novel resolution to a typical fuzzy transportation problem

was originated by [15]. [21] analyzed symmetric fuzzy numbers, specified an equivalent

definition of convex fuzzy sets, and proposed a technique for creating a symmetric con-

vex fuzzy set. [17] proclaimed that no underlying membership functions were needed

and explained many aggregation operators. A further procedure for defuzzifying gen-

eralized fuzzy numbers was designed by [22]. [18] presented and approximated many

strategies for total ordering in the intuitionistic fuzzy numbers (IFNs) category. [10]

suggested ranking fuzzy numbers with widely deviating stretches. [13] presented an

alternate technique for estimating additional arithmetic operations of a system using

the sigmoidal number under the fuzzy background. [23] involved fuzzy arithmetic oper-

ations to equations employed to describe anticipation in multiple applications. The

idea of pentagonal-fuzzy number (PFN) was submitted in a broad insight by [20].[6]

formulated pentagonal-fuzzy numbers with other expressions to analyze the conse-

quence of optimal strategies for the candidates in real game situations.

This article covers all defuzzification formulas, expressions, and techniques by consid-

ering the continuity of membership functions. The solution to the game problem is

ultimately detected, and the results are compared to [6]. Subsequently, the present This article is prepared: Section 2 includes preliminaries and pentagonal-fuzzy num-paper provides the updated version of optimal strategies and values of matrix games bers with various expressions, and their α cuts are consulted in Section 3. Section mentioned in [6].

4 contracts with three defuzzification procedures of linear pentagonal fuzzy numbers

with symmetry. In Section 5 hierarchy of pentagonal-fuzzy numbers is depicted. Three

numerical illustrations and suggested defuzzification approaches have numerically

resembled in Section 6. Section 7 finishes with the conclusion of this paper.

2. Preliminaries

In this section, we recall some basic definitions and notations which are useful

throughout the paper.

Definition 1. A fuzzy set A ˜ is the collection of ordered pairs (x, µ (x)) defined on

the universal set X and denoted by A ˜ = {(x, µ (x))|x ∈ X and µ (x) ∈ [0, 1]}, where

A ˜ A ˜

µ : X → [0, 1] is the membership function of A ˜It is commonly believed that all

details accessed by the entities are represented in crisp numbers. However, most real-

life systems are engaged when the goals and constraints are often vague or unclear.

Definition 2. An interval valued fuzzy set A ˜ defined on the universal set X is

represented by A ˜ = {(x, (µ (x), µ (x))) |x ∈ X}, where µ : X → [0, λ] and

A ˜U A ˜L A ˜U µ ˜L : X → [0, ω] are the upper and lower membership functions with the maximum

values λ and ω respectively.

Definition 3. A non linear interval valued triangular fuzzy number A ˜ is

nLIV F N

denoted by A ˜ = [{(a , b, c ; λ) , (a, b, c; ω)}; ξ , ξ , ξ , ξ ] where 0 < ω ≤ λ ≤ 1

nLIV F N 1 1 1 2 3 4

and a1 < a < b < c < c1. The upper and lower membership functions are defined as

follows:  ξ

 x−a1  λ ; a1 ≤ x ≤ b  b−a1 λ ; x = b µ ˜U (x) = (1)

A ξ2

 c1−x  λ ; b ≤ x ≤ c  c −b 1 0 ; else

and   x−a  ω ; a ≤ x ≤ b  b−a ω ; x = b µ ˜L (x) = (2)

A ξ4

 c−x  ω ; b ≤ x ≤ c  c−b 0 ; else

3. Pentagonal fuzzy number and its different

representations

Within this section, we have established distinct categories of pentagonal fuzzy num-

bers by leveraging the continuity of the membership function—a dimension that was

notably absent in the framework presented by [6].

Definition 4. A pentagonal fuzzy number with µ ˜(x) as membership function is

represented by A ˜ = (a , a , a , a , a ) and satisfies the following conditions:

1 2 3 4 5

1. µ ˜(x) is continuous in [0, 1].

2. µ ˜(x) is strictly increasing continuous function on [a1, a2] and [a2, a3].

3. µ ˜(x) is strictly decreasing continuous function on [a3, a4] and [a4, a5].

Definition 5. Linear pentagonal fuzzy number with symmetry (LS) is represented by

A ˜ = ⟨a , a , a , a , a ; k⟩. The revised definition by the Figure 1 with a continuous

LS 1 2 3 4 5

membership function is given  as follows:

 x−a1  k ; a1 ≤ x ≤ a2  a2−a1  x−a  k + (1 − k) ; a2 ≤ x ≤ a3  a3−a2 1 ; x = a3 µ ˜ (x) = (3)

ALS a −x

 k + (1 − k) ; a3 ≤ x ≤ a4  a4−a3  a −x  k ; a4 ≤ x ≤ a5  a5−a4 0 ; else Fig. 1 Linear PFN with symmetry (LS).

Definition 6. The revised α-cut of linear pentagonal fuzzy number with symmetry is defined as the collection of all x ∈ X having membership function µ ˜ (x) as greater

A LS

than or equal to α i.e. Aα = {x ∈ X|µ ˜ (x) ≥ α} is given by

ALS

 A1L(α) = a1 + (a2 − a1) ; α ∈ [0, k]  α−k A2L(α) = a2 + (a3 − a2) ; α ∈ [k, 1] Aα = α−k (4)  A2R(α) = a4 − (a4 − a3) ; α ∈ [k, 1]  1−k A1R(α) = a5 − (a5 − a4) ; α ∈ [0, k]

Here A1L(α), A2L(α) are the increasing functions of α and A1R(α), A2R(α) are the

decreasing functions of α.

Definition 7. Non linear pentagonal fuzzy number with symmetry (N S) is represented

by A ˜ = (a , a , a , a , a ; k) . The revised definition by the Figure 2 with

N S 1 2 3 4 5 (ξ ,ξ ;ζ ,ζ )

1 2 1 2

a continuous membership function is given as follows:

 x−a1  k ; a1 ≤ x ≤ a2  a2−a1  ξ2  x−a  k + (1 − k) ; a2 ≤ x ≤ a3  a3−a2 1 ; x = a3 µ ˜ (x) = ζ (5)

AN S 1

 a4−x  k + (1 − k) ; a ≤ x ≤ a  a −a 3 4  ζ2  a5−x  k ; a4 ≤ x ≤ a5  a −a  5 4 0 ; else Fig. 2 Non Linear PFN with Symmetry A ˜

NS.

Definition 8. The revised α-cut of non linear pentagonal fuzzy number with symmetry

is defined as the collection of all x ∈ X with membership function µ ˜ (x) as greater

AN S

than or equal to α i.e. Aα = {x ∈ X|µ ˜ (x) ≥ α} is

AN S

 α k (a2 − a1) ; α ∈ [0, k]  A1L(α) = a1 + 1

 α−k (a − a ) ; α ∈ [k, 1]  A2L(α) = a2 + 1−k 3 2  (6) Aα = α−k  A (α) = a − 1−k (a4 − a3) ; α ∈ [k, 1]  2R 4  α 2  A (α) = a − k (a5 − a4) ; α ∈ [0, k]  1R 5

Here A1L(α), A2L(α) are the increasing functions of α and A1R(α), A2R(α) are the

decreasing functions of α.

Definition 9. The revised definition of linear interval valued pentagonal fuzzy num-

ber with symmetry (LIP S) A ˜ = ⟨(a , a , c, a , a ; k, p) , (b , b , c, b , b ; w, q)⟩ is

LIP S 1 2 4 5 1 2 4 5 defined by the Figure 3 with upper and lower membership functions as follows:

 x−a1  p ; a1 ≤ x ≤ a2  a2−a1  x−a  p + (k − p) ; a2 ≤ x ≤ c  c−a2 k ; x = c µ ˜U (x) = (7)

A a −x

LIP S  4  p + (k − p) ; c ≤ x ≤ a4  a4−c  a −x  p ; a4 ≤ x ≤ a5  a5−a4 0 ; else

and   x−b1  q ; b1 ≤ x ≤ b2  b2−b1  x−b  q + (w − q) ; b2 ≤ x ≤ c  c−b2 w ; x = c µ ˜L (x) = (8)

A b −x

LIP S  4  q + (w − q) ; c ≤ x ≤ b4  b4−c  b −x  q ; b4 ≤ x ≤ b5  b5−b4 0 ; else

Figure 3
Figure 3. Linear IVPFN with symmetry (LIPS).

Definition 10. The revised α-cut of linear interval valued pentagonal fuzzy number

˜ ˜U ˜L with symmetry is defined as ALIP Sα = ALIP S ⊖ ALIP S for α1 ∈ [0, w];

α α2 α α1 α2 ∈ [0, k].

Where  U α2 (a − a ) ; α ∈ [0, p]  A (α ) = a + p 2 1 2  1L 2 1  U α2−p  A (α ) = a + (c − a2) ; α2 ∈ [p, k] 2L 2 2 k−p (9) ˜U U α2−p A = A (α ) = a − (a4 − c) ; α2 ∈ [p, k]

LIP S 2R 2 4 k−p

α2  α  U 2 (a − a ) ; α ∈ [0, p]  A 1R(α2) = a5 − p 5 4 2

and  L α1 (b − b ) ; α ∈ [0, q]  A (α ) = b + q 2 1 1  1L 1 1  L α1−q  A (α ) = b + (c − b2) ; α1 ∈ [q, w] 2L 1 2 w−q (10) ˜L L α1−q A = A (α ) = b − (b4 − c) ; α1 ∈ [q, w]

LIP S 2R 1 4 w−q

α1  α  L 1 (b − b ) ; α ∈ [0, q]  A 1R(α1) = b5 − q 5 4 1

U U L L

Here A 1L(α2), A 2L(α2), A 1L(α1), A 2L(α1) are the increasing functions of

U U L L

α2 and α1 correspondingly and A 1R(α2), A 2R(α2), A 1R(α1), A 2R(α1) are the

decreasing functions of α2 and α1 correspondingly.

Definition 11. The membership functions of linear interval val-

ued pentagonal fuzzy number with asymmetry (LIP AS) A ˜ =

LIP AS

⟨(a1, a2, c, a4, a5; k, p, r) , (b1, b2, c, b4, b5; w, q, s)⟩ are defined by the Figure 4 with upper

and lower membership functions as follows:

; a1 ≤ x ≤ a2  a2−a1  x−a1  p x−a2 ; a ≤ x ≤ c  p + (k − p) c−a 2

 k ; x = c (11)

 a4−x ; c ≤ x ≤ a4 r + (k − r) a4−c µ ˜U (x) = a −x A 5 ; a ≤ x ≤ a LIP AS r a −a 4 5  5 4

 0 ; else

and  x−b ; b1 ≤ x ≤ b2  b2−b1 1  x−b2  q + (w − q) ; b2 ≤ x ≤ c  c−b2  w ; x = c (12)

b4−x ; c ≤ x ≤ b4 µ ˜L (x) = s + (w − s) b4−c LIP AS   b5−x ; b ≤ x ≤ b  s b −b 4 5  5 4  0 ; else

Definition 12. The revised α-cut of linear interval valued pentagonal fuzzy number

˜U ˜L with asymmetry is defined as ALIP ASα = ALIP AS ⊖ ALIP AS for α1 ∈

α α2 α α1

[0, w]; α2 ∈ [0, k].

Figure 4
Figure 4. Linear IVPFN with asymmetry (LIPAS).

Where 

U α2

A (α ) = a + (a2 − a1) ; α2 ∈ [0, p]  1L 2 1 p  α −p  U 2 A (α ) = a + (c − a2) ; α2 ∈ [p, k] U 2L 2 2 k−p (13) A ˜ = U α −r LIP AS 2 (a − c) ; α ∈ [r, k] A 2R(α2) = a4 − k−r 4 2 α2   U α  2 (a − a ) ; α ∈ [0, r] A 1R(α2) = a5 − r 5 4 2

and  L α1 (b − b ) ; α ∈ [0, q] A (α ) = b + q 2 1 1  1L 1 1  L α1−q  A (α ) = b + (c − b2) ; α1 ∈ [q, w] 2L 1 2 w−q (14) ˜L α −s A = L 1 (b − c) ; α ∈ [s, w] LIP AS A 2R(α1) = b4 − w−s 4 1 α1   L α1  A (α ) = b − (b5 − b4) ; α1 ∈ [0, s]

1R 1 5 s U U L L

Here A 1L(α2), A 2L(α2), A 1L(α1), A 2L(α1) are the increasing functions of

U U L L

α2 and α1 correspondingly and A 1R(α2), A 2R(α2), A 1R(α1), A 2R(α1) are the

decreasing functions of α2 and α1 correspondingly.

Definition 13. The membership functions of non linear interval val-

ued pentagonal fuzzy number with symmetry (N IP S) A ˜ = D E N IP S

(a1, a2, c, a4, a5; k, p) , (b1, b2, c, b4, b5; w, q) are defined by the Figures 5 to 8 for different (ξ1,ξ2;ζparameters 1,ζ2) ξ1, ξ2, ζ1, ζ2 with (ξ1,ξrevised 2;ζ1,ζ2) continuous upper and

lower membership functions as follows:

Figure 5
Figure 5. A ˜ for ξ , ξ , ζ , ζ > 1.

NIPS 1 2 1 2

 x−a1 ξ1  ; a1 ≤ x ≤ a2  a2−a1

x−a2 ; a ≤ x ≤ c  p + (k − p) c−a 2 k ; x = c (15) µ ˜U (x) = ζ1 N IP S   a4−x ; c ≤ x ≤ a4  p + (k − p) a4−c

 a5−x ; a ≤ x ≤ a  p a −a 4 5

 0 ; else

and ξ1 ; b1 ≤ x ≤ b2  bx 2− −bb1 1  q ξ x−b2 ; b ≤ x ≤ c  q + (w − q) c−b 2  w ; x = c (16)  ζ1 b4−x ; c ≤ x ≤ b4 µ ˜L (x) = q + (w − q) b4−c N IP S  2 b5−x ; b ≤ x ≤ b  q b −b 4 5  5 4  0 ; else

Definition 14. The revised α-cut of non linear interval valued pentagonal fuzzy num-

˜U ˜L ber with symmetry is defined as follow: AN IP Sα = AN IP S ⊖ AN IP S for

α α2 α α1

α1 ∈ [0, w]; α2 ∈ [0, k].

Figure 6
Figure 6. A ˜ for ξ , ξ > 1; ζ , ζ < 1.

NIPS 1 2 1 2

Figure 7
Figure 7. A ˜ for ξ , ξ , ζ , ζ < 1.

NIPS 1 2 1 2 NIPS 1 2 1 2

Figure 8
Figure 8. A ˜ for ξ , ξ < 1; ζ , ζ > 1.

Where

 U α2  A (α ) = a + p (a2 − a1) ; α2 ∈ [0, p]  1L 2 1  ξ2  U α −p  2 (c − a ) ; α ∈ [p, k]  A 2L(α2) = a2 + k−p 2 2 (17) A = ζ1

N IP S U α2−p

α  A (α ) = a − k−p (a4 − c) ; α2 ∈ [p, k] 2  2R 2 4

 U α2  A (α ) = a − p (a5 − a4) ; α2 ∈ [0, p]  1R 2 5

and 1  L α1 (b2 − b1) ; α1 ∈ [0, q]  A (α ) = b + q  1L 1 1 1  α −q  L 1 (c − b ) ; α ∈ [q, w]  A 2L(α1) = b2 + w−q 2 1 (18) ˜L ζ1 A = L α1−q N IP S A (α ) = b − w−q (b4 − c) ; α1 ∈ [q, w] α1  2R 1 4  ζ2  L α  1 (b − b ) ; α ∈ [0, q]  A 1R(α1) = b5 − q 5 4 1

U U L L

Here A 1L(α2), A 2L(α2), A 1L(α1), A 2L(α1) are the increasing functions of

U U L L

α2 and α1 correspondingly and A 1R(α2), A 2R(α2), A 1R(α1), A 2R(α1) are the

decreasing functions of α2 and α1 correspondingly.

Definition 15. The membership functions of non linear interval val-

ued pentagonal fuzzy number with asymmetry (N IP AS) A ˜ =

N IP AS (a1, a2, c, a4, a5; k, p, s) , (b1, b2, c, b4, b5; w, q, t) are defined by

(ξ1,ξ2;ζ1,ζ2) (ξ1,ξ2;ζ1,ζ2)

the Figures 9 to 12 for different parameters ξ1, ξ2, ζ1, ζ2 having continuous upper and

lower membership functions as follows:

 x−a1 ξ1  ; a1 ≤ x ≤ a2  a2−a1

x−a2 ; a ≤ x ≤ c  p + (k − p) c−a 2 k ; x = c (19) µ ˜U (x) = ζ1 N IP AS   a4−x ; c ≤ x ≤ a4  s + (k − s) a4−c

a5−x ; a ≤ x ≤ a  s a −a 4 5  5 4  0 ; else

and ξ1 ; b1 ≤ x ≤ b2  bx 2− −bb1 1  q ξ2  x−b2 ; b ≤ x ≤ c  q + (w − q) c−b 2  w ; x = c (20)

b4−x ; c ≤ x ≤ b4 µ ˜L (x) = t + (w − t) b4−c N IP AS  2  b −x  5 ; b ≤ x ≤ b  t b −b 4 5  5 4  0 ; else

Figure 9
Figure 9. A ˜ for ξ , ξ , ζ , ζ > 1.

NIPAS 1 2 1 2 NIPAS 1 2 1 2

Figure 10
Figure 10. A ˜ for ξ , ξ > 1; ζ , ζ < 1.

Definition 16. The revised α-cut of non linear interval valued pentagonal fuzzy num-

˜U ˜L ber with asymmetry defined as AN IP ASα = AN IP AS ⊖ AN IP AS for

α α2 α α1

α1 ∈ [0, w]; α2 ∈ [0, k].

Where

 α p (a2 − a1) ; α2 ∈ [0, p]  U 2  A 1L(α2) = a1 + 1  U α2−p  A (α ) = a + k−p (c − a2) ; α2 ∈ [p, k] (21)  2L 2 2 ˜U ζ A = 1 N IP AS U α2−s (a − c) ; α ∈ [s, k] α  A (α ) = a − k−s 4 2 2  2R 2 4  U α2 2  A (α ) = a − s (a5 − a4) ; α2 ∈ [0, s]  1R 2 5

and 1

 L α1 (b − b ) ; α ∈ [0, q]  A (α ) = b + q 2 1 1  1L 1 1 1  ξ2  L α1−q  A (α ) = b + w−q (c − b2) ; α1 ∈ [q, w] (22)

2L 1 2

A ˜ =

N IP AS ζ1

α  L α1−t (b − c) ; α ∈ [t, w] 1  A (α ) = b − w−t 4 1  2R 1 4  L α1 2  A (α ) = b − t (b5 − b4) ; α1 ∈ [0, t]  1R 1 5 NIPAS 1 2 1 2

Figure 11
Figure 11. A ˜ for ξ , ξ , ζ , ζ < 1.
Figure 12
Figure 12. A ˜ for ξ , ξ < 1; ζ , ζ > 1.

NIPAS 1 2 1 2 U U L L

Here A 1L(α2), A 2L(α2), A 1L(α1), A 2L(α1)”are the increasing functions of

U U L L

α2 and α1 correspondingly and A 1R(α2), A 2R(α2), A 1R(α1), A 2R(α1) are the

decreasing functions of α2 and α1 correspondingly.

4. Defuzzification

The defuzzification method converts the fuzzy inference result into an appropriate

exact value. It plays a critical role in a fuzzy environment. For fuzzy logic, several

defuzzification techniques are available, such as the center of the area (centroid), great-

est of maxima, smallest of maxima, mean of maxima, the bisector of area, graded

mean integral value, and so on. We investigate three methods to defuzzify the linear

pentagonal fuzzy number with symmetry (LS) based on different approaches. Our

results are compared with [6].

4.1 Centroid Method

In this section we have proposed R a methodR for the defuzzification R of non linear

a2 a3 a4 a5

xµ ˜ (x)dx+ xµ ˜ (x)dx+ xµ ˜ (x)dx+ xµ ˜ (x)dx a1 R AN S a R2 AN S Ra3 AN S R a4 AN S

a2 a3 a4 a5

pentagonal fuzzy number µ (with x)dx+symmetry µ (x) as dx+ follows: µ (x)dx+ µ (x)dx (23)

a A ˜ a A ˜ a A ˜ a A ˜ 1 N S 2 N S 3 N S 4 N S

= R Q

From Eqs. (5) and (23) we obtain " # Z " # Z ξ1 a ξ2

a2 3

x − a1 x − a2 P = x k dx + x k + (1 − k) dx a2 − a1 a a3 − a2

a1 2

# Z " # " ζ1 a ζ2 a4 a5 − x a4 − x dx + x k dx + x k + (1 − k) a − a a − a

4 3 a4 5 4

# Z " # and " ξ1 a3 ξ2 Z a x − a2 x − a1 dx + k + (1 − k) dx Q = k a2 − a1 a a3 − a2 # Z " # a1 " ζ ζ

1 a5 2

Z a a − x a4 − x dx + k dx + k + (1 − k) a − a a − a

4 3 a4 5 4

On computing we have

2 2 2

k(a −a ) k(a3 −a2 ) (1−k)a (a −a )

2 1 3 3 2

ka (a −a ) − + +

2 (ξ1+1) 2 1 (ξ1+1)(ξ1+2) 2 (ξ2+1)

P = 2 2 2 k(a4 −a3 ) (1−k)a (a −a ) (24)

(1−k)(a −a ) 3 4 3

3 2 + + − (ξ2+1)(ξ2+2) 2 (ζ1+1)

2 ka4(a5−a4) k(a5−a4) (1−k)(a4−a3)

+ + +

(ζ1+1)(ζ1+2) (ζ2+1) (ζ2+1)(ζ2+2) and

(1−k)(a3−a2) (1−k)(a4−a3)

k(a2−a1) + k (a − a ) + + k (a − a ) + Q = (ξ +1) 3 2 (ξ +1) 4 3 (ζ +1) 1 2 1 (25)

k(a5−a4)

+ (ζ +1)

If we set ξ1 = 1, ξ2 = 1, ζ1 = 1 and ζ2 = 1 then the proposed defuzzification technique

applies for the linear pentagonal fuzzy number with symmetry as follows

(26) L + NM

where

2 2 2 2

k(a −a ) k(a3 −a2 ) (1−k)a (a −a ) (1−k)(a −a ) 2 1 3 3 2 3 2 (27) − + + −

ka2(a2 2−a1) 6 2 2 6

2 2 (1−k)a3(a4−a3) (1−k)(a4−a3) ka4(a5−a4) k(a5−a4)

k(a4 −a3 ) (28) + + + + M = 2 2 6 2 6 and (1−k)(a −a ) k(a2−a1) + k (a − a ) + + k (a − a ) N = 2 3 2 2 4 3 (29)

k(a5−a4) (1−k)(a4−a3)

For k = 1 we have

2 2 2 2 (30) a4 + a5 + a4a5 − a1 − a2 − a1a2 R = 3 (a4 + a5 − a1 − a2)

4.2 Defuzzification using α-cut

In this section, we propose a method to compute the defuzzification of non-linear

pentagonal D fuzzy numbers with symmetryE as follows:

Since try A ˜ the= left(a and , aright , a , a α-cuts , a ; kof ) a non linear are pentagonal fuzzy number with symme-

N S 1 2 3 4 5 (ξ ,ξ ;ζ ,ζ )

1 2 1 2

−1 α (a − a ); α ∈ [0, k], L (α) = a1 + k

−1 2

R (α) = a5 − (a5 − a4); α ∈ [0, k],

α − k ξ2 L (α) = a2 + (a3 − a2); α ∈ [k,1], 1 − k

α − k ζ1 R (α) = a4 − (a4 − a3); α ∈ [k,1].

1 − k Then the mean of α- cut of A ˜ is

Z 1 −1 −1 L (α) + R (α) dα A ˆ = 2 Z 1 −1 −1 Z kα=0 L (α) + R (α)

−1 −1

L (α) + R (α) dα + dα = 2 2 α=k i α=0 h 1 1

ξ1 α ζ2

Z α (a − a ) + a − (a − a ) k a1 + k 2 1 5 k 5 4 dα

ξ2 ζ1

α−k α−k

Z a + 1−k (a3 − a2) + a4 − 1−k (a4 − a3) dα

α=k k(a5−a4) (1−k)(a3−a2) (1−k)(a4−a3)

k(a2−a1) − + (1 − k) (a + a ) + − k (a + a ) + 2 4

1 5 1 +1 1 +1 1 +1 1 +1

ξ1 ζ2 ξ2 ζ1

(31)

If we set ξ1 = 1, ξ2 = 1, ζ1 = 1 and ζ2 = 1 then the proposed defuzzification technique

applies for the linear pentagonal fuzzy number with symmetry as follows

A ˆ = k(a −a ) (1−k)(a −a ) (1−k)(a −a )

5 4 3 2 4 3

− + (1 − k) (a2 + a4) + −

2 2 2 2 k(a2−a1)

k (a1 + a5) + 2 (32)

For k = 1 we have a1 + a2 + a4 + a5 (33) A ˆ = 4

4.3 Defuzzification using Removal of Area Method

We consider numerous types of regions of the related linear PFN 1 as illustrated below R A, ˆ 0 = Area of the highlighted region for Figure 13 = k (a + a )

1 2 1 2

R2 A, 0 = Area of the highlighted region for Figure 14 = (1 − k) (a2 + a3)

R3 A, 0 = Area of the highlighted region for Figure 15 = (1 − k) (a3 + a4)

R4 A, 0 = Area of the highlighted region for Figure 16 = k (a4 + a5)

Therefore,

R A, ˆ 0 + R A, ˆ 0 + R A, ˆ 0 + R A, ˆ 0

1 2 3 4

R D, ˆ 0 = 4 1 1 1 (34) 1 k (a + a ) + (1 − k) (a + a ) + (1 − k) (a + a ) + k (a + a )

2 1 2 2 2 3 2 3 4 2 4 5

k (a1 + a2) + (1 − k) (a2 + a3) + (1 − k) (a3 + a4) + k (a4 + a5).

Figure 13
Figure 13. First step for removal of region method.

For k = 1, we have a1 + a2 + a4 + a5 R D, ˆ 0 = . (35)

4.4 Comparison of the Above Defuzzification Methods

We compared the above proposed defuzzification methods numerically for suitable

pentagonal fuzzy number.

Table 1. Numerically Comparison of Defuzzification Methods.

This table is shown in the PDF version of the article.

5. Numerical Problems

Example 1. Consider a fuzzy matrix game for two different players A and B whose

pay-offs are linear pentagonal fuzzy numbers with symmetry as follows:

Now using the mean of α- cut method as defuzzification defined by Eq. (32)

for linear pentagonal fuzzy number to convert the given fuzzy pay-off matrix game as Table (1) into the following crisp matrix game as Table 3, we have

Figure 14
Figure 14. Second step for removal of region method.
Figure 15
Figure 15. Third step for removal of region method.

The above crisp matrix game defined by Table 3 has no saddle point. Therefore according to the dominance property in matrix games row X1 can be dominated by row X3. Hence we

have a modified matrix game as Table 4.

Again we have that all the elements of column Y1 are greater than or equal to corresponding elements of column Y3. Therefore column Y1 dominates column Y3 and so we can delete the

column Y1 for new improved matrix game as Table 5.

Figure 16
Figure 16. Fourth step for removal of region method
Table 3. Defuzzified pay-off matrix game.

This table is shown in the PDF version of the article.

Table 4. First step of dominance property.

This table is shown in the PDF version of the article.

Now each element of average of strategies Y3 and Y4 of player B are less than or

equal to the corresponding elements of strategy Y2 of player B. Therefore on deleting

the column of strategy Y2 by dominance property, we obtain a new modified matrix

game as Table (6).

Again each element of strategy X2 of player A is less than or equal to the corresponding

elements of the average of strategies X3 and X4 of player A. Therefore on deleting the

row of strategy X2 by dominance property, we obtain a new improved matrix game

as Table (7).

2 1 2 1 8

Which are A 0 is, 0 a,2× , 2 matrix and B game 0, 0, without , respectively and the value of game is saddle point, so the strategy of both.the players

3 3 3 3 3 Example 2. Consider a fuzzy matrix game for two different players A and B whose pay-offs are linear pentagonal fuzzy numbers with symmetry as follows:

Table 5. Second step of dominance property.

This table is shown in the PDF version of the article.

Table 6. Third step of dominance property.

This table is shown in the PDF version of the article.

Table 7. Fourth step of dominance property.

This table is shown in the PDF version of the article.

Now using the mean of α-cut method as defuzzification defined by equation (32) for the linear pentagonal fuzzy number to convert the given fuzzy pay-off matrix game

as Table (8) into the crisp matrix game defined by the Table (9), as follows:

Table 8. Defuzzified pay-off matrix game.

This table is shown in the PDF version of the article.

Since the above crisp matrix game defined by Table (8) has no saddle point, there-

fore according to the dominance property in matrix games, strategy X1 of player A

can be dominated by strategy X3 of player A. Hence we have a modified matrix game

defined by Table (9).

Table 9. First step of dominance property.

This table is shown in the PDF version of the article.

Again we have that all the elements of strategy Y1 are greater than or equal to the

corresponding elements of strategy Y3 for player B. Therefore strategy Y1 dominates

strategy Y3 and so we can delete the strategy column Y1 for new improved matrix

game as Table (10).

Table 10 Second step of dominance property.

Strategy Y2 Y3 Y4

X2 5 3 5 X3 3 5 0 X4 4 0 9

Further the matrix game as Table (10) can not be reduced by dominance rule. There-

fore by applying the linear programming approach for matrix game, we have

For Player A 

 max v  such that  5x1 + 3x2 + 4x3 ≥ v 3x1 + 5x2 + 0x3 ≥ v  5x1 + 0x2 + 9x3 ≥ v  x + x + x = 1  1 2 3 and x1, x2, x3, v ≥ 0.

and

For Player B  min w

 such that  5y + 3y + 5y ≤ w  1 2 3 3y1 + 5y2 + 0y3 ≤ w  4y + 0y + 9y ≤ w  1 2 3  y1 + y2 + y3 = 1 and y , y , y , w ≥ 0.

1 2 3

Where v, w are the minimum expected gain and maximum expected loss for player A

and B respectively. The above linear programming problems for player A and player B as x y can be written in the following manner by assuming = Ri, = Sj for i, j = 1, 2, 3 For Player A  v w  min R1 + R2 + R3

 such that

5. R1 + 3R2 + 4R3 ≥ 1

 3R1 + 5R2 + 0R3 ≥ 1  5R1 + 0R2 + 9R3 ≥ 1

and R1, R2, R3 ≥ 0.

and

For Player B  max S1 + S2 + S3

 such that  5S + 3S + 5S ≤ 1

1 2 3

3. S1 + 5S2 + 0S3 ≤ 1

 4S + 0S + 9S ≤ 1  1 2 3 and S1, S2, S3 ≥ 0.

On solving the above linear programming problems by the classical simplex method, 0.28. So the optimal strategy of both the players A and B are A (0, 0.1 71429, 0.28571 1, 0) we obtain R1 = 0.2; R2 = 0.08; R3 = 0; S1 = 0; S2 = 0.2; S3 = 0.08; = 0.28 and = and B (0, 0, 0.71429, 0.28571) respectively and the value of game is 3.57143. In [6] the

9 5 9 5

optimal strategies of the players A 0, 0, , and B 0, 0, , and the value of

14 14 14 14

45 were evaluated.

game 14

Example 3. Consider a fuzzy matrix game for two different players A and B whose pay-offs are linear pentagonal fuzzy numbers with symmetry as follows:

Now using the mean of α- cut method as defuzzification defined by Eq. (32) for the linear pentagonal fuzzy number to convert the given fuzzy pay-off matrix game

as Table (11) into the crisp matrix game defined by the Table (13), as follows: Since the above crisp matrix game defined by Table (11) has no saddle point,

Table 11. Defuzzified pay-off matrix game.

This table is shown in the PDF version of the article.

Table 12. Defuzzified pay-off matrix game obtained by [6].

This table is shown in the PDF version of the article.

therefore according to the dominance property in matrix games, strategy X1 can be

dominated by strategy X3 of player A. Hence we have a modified matrix game defined

by Table (13).

Table 13. First step of dominance property.

This table is shown in the PDF version of the article.

Again we have that all the elements of strategy Y1 are equal to the corresponding

elements of strategy Y3 for player B. Therefore either strategy Y1 or strategy Y3 for

player B can be dominated by each other. Here we delete the strategy Y1 for the new,

improved matrix game as Table (14).

Now each element of the average of strategies Y3 and Y4 of player B are less than or

equal to the corresponding elements of strategy Y2 of player B. Therefore on deleting

the column strategy Y2 by dominance property, we obtain a newly modified matrix

game defined by the following Table (15) as Further the matrix game as Table (14) can not be reduced by dominance rule.

Table 14. Second step of dominance property.

This table is shown in the PDF version of the article.

Table 15. Third step of dominance property.

This table is shown in the PDF version of the article.

Therefore by applying the graphical approach for m × 2 matrix game according as

Table (15) given by Figure 17.

Hence it is clear that the lowest boundary point P on the bounded region will give

the smallest expected loss to the minimizing player (playerB). So the best strategy

for player A are those which pass through the point P. Therefore we can write the

matrix game as given by the Table (15) in a 2 × 2 matrix game defined by the Table

(16) as follows:

Table 16. Fourth step of dominance using graphical method.

This table is shown in the PDF version of the article.

The above 2 × 2 matrix game is without saddle point, so the strategy of both the

6 1 6 1 36

players are A 0, , , 0 and B 0, 0, , respectively and the value of game is . In

7 7 7 7 7 5 3 5 3

[6] the optimal strategies of the players A 0, 0, , , B 0, 0, , and the value of

8 8 8 8

the game were evaluated.

Remark 1. If we delete the strategy Y3 for new improved matrix game as second step

of dominance property and further on going in a similar manner, we obtain the best

1 6 1

6 , , 0 and B , 0, 0, respectively and the strategy for player A and B as A 0, 7 7 7 7 value of game is .

5.1 Comparative Study of Proposed Defuzzification Techniques

by Numerical Examples

In this section we compared all the proposed defuzzification techniques numerically

for matrix game problems assigned by [6] and investigate that I. The strategy of the game players remains unchanged in realistic game problems.

Figure 17
Figure 17. Graphical representation for third step of dominance property.

II. Centroid method and mean of α-cut method provides the same results.

III. Value of the matrix game obtained by removal of region method is the half of the

game value obtained by other remaining proposed methods.

6. Conclusion

This research paper has systematically explored and elucidated a range of mathemat-

ical concepts central to the pentagonal fuzzy numbers. The investigation developed

various concepts, including distinct representations, intrinsic properties, α-cuts, and defuzzification methodologies. The application of both linear and non-linear member-ship functions in a symmetric fashion has been rigorously examined, contributing to a comprehensive understanding of the subject. The incorporation of numerical exam-ples drawn from matrix game problems, as originally provided by [6], has added depth and context to the discourse. The introduction and analysis of novel defuzzification techniques for pentagonal fuzzy numbers have provided valuable insights into optimal strategies for players, as well as the determination of game values—parameters of sub-stantial significance within real-world gaming scenarios. This paper offers a stepping stone for further scholarly inquiries into the field of pentagonal fuzzy numbers across diverse real-life problem domains. By shedding light on the representation, defuzzifica-tion, and their applications within the context of game theory, this research develops the way for continued exploration and innovative utilization of pentagonal fuzzy num-bers in addressing complex phenomena. Researchers are encouraged to build upon this foundation, unraveling the potential of pentagonal fuzzy numbers as a powerful tool to handle the real-world challenges.

Acknowledgments

The authors would like to express their appreciation to all individuals whose valuable comments and suggestions contributed to the improvement of this research.

Funding

The authors declare that no financial support, grant, or sponsorship was received for conducting this study.

Data Availability

The data used and analyzed during the current study are available from the corre-sponding author upon reasonable request.

Figures

Figure
Figure. Image from the PDF file; its caption could not be matched.
Figure
Figure. Image from the PDF file; its caption could not be matched.

References

  1. [1] Abbasbandy, S., & Hajjari, T. (2009). A new approach for ranking of trapezoidal fuzzy numbers. Computers & mathematics with applications, 57(3), 413-419. https://doi.org/10.1016/j.camwa.2008.10.090
  2. [2] Asady, B., & Zendehnam, A. (2007). Ranking fuzzy numbers by distance mini-mization. Applied mathematical modelling, 31(11), 2589- 2598.https://doi.org/10.1016/j.apm.2006.10.018
  3. [3] Atanassov, K. T. (1999). Intuitionistic fuzzy sets. In Intuitionistic fuzzy sets: theory and applications (pp. 1-137). Heidelberg: Physica-Verlag HD. https://doi.org/10.1007/978-3-7908-1870-3_1
  4. [4] Atanassov, K. T. (1986) ‘Intuitionistic fuzzy sets’, Fuzzy Sets Syst, 20(1), 87–96.
  5. [5] Buckley, J. J. (1988). Possibilistic linear programming with triangular fuzzy numbers. Fuzzy sets and Systems, 26(1), 135-138. https://doi.org/10.1016/0165-0114 (88)90013-9
  6. [6] Chakraborty, A., Mondal, S. P., Alam, S., Ahmadian, A., Senu, N., De, D., & Salahshour, S. (2019). The pentagonal fuzzy number: its different representa-tions, properties, ranking, defuzzification and application in game problems. Symmetry, 11(2), 248.
  7. [7] Chen, S. M. (2011). Analyzing fuzzy risk based on a new fuzzy ranking method between generalized fuzzy numbers. Expert Systems with Applications, 38(3), 2163- 2171. https://doi.org/10.1016/j.eswa.2010.08.002
  8. [8] Cheng, C. H. (1998). A new approach for ranking fuzzy numbers by distance method. Fuzzy sets and systems, 95(3), 307-317. https://doi.org/10.1016/S0165-0114 (96)00272-2
  9. [9] Chou, C. C. (2003). The canonical representation of multiplication operation on triangular fuzzy numbers. Computers & Mathematics with Applications, 45(10-11), 1601-1610. https://doi.org/10.1016/S0898-1221(03)00139-1
  10. [10] Chutia, R., & Chutia, B. (2017). A new method of ranking parametric form of fuzzy numbers using value and ambiguity. Applied soft computing, 52, 1154- 1168. https://doi.org/10.1016/j.asoc.2016.09.013
  11. [11] Dubois, D., & Prade, H. (1978). Operations on fuzzy numbers. International jour-nal of systems science, 9(6), 613-626. https://doi.org/10.1080/00207727808941724
  12. [12] Dubois, D., & Prade, H. (Eds.). (2012). Fundamentals of fuzzy sets (Vol. 7). Springer Science & Business Media. https://link.springer.com/book/10.1007/978-1-4615 -4429-6
  13. [13] Garg, H. (2018). Some arithmetic operations on the generalized sigmoidal fuzzy numbers and its application. Granular computing, 3(1), 9-25.. https://doi.org/10.1007/s41066-017-0052-7
  14. [14] Heilpern, S. (1997). Representation and application of fuzzy numbers. Fuzzy sets and systems, 91(2), 259-268. https://doi.org/10.1016/S0165-0114(97)00146-2
  15. [15] Kaur, A., & Kumar, A. (2012). A new approach for solving fuzzy transportation prob-lems using ´ generalized trapezoidal fuzzy numbers. Applied soft computing, 12(3), 1201- Kosiń ski, W., Prokopowicz, P., & Ś lę zak, D. (2003). Ordered fuzzy numbers. 1213. https://doi.org/10.1016/j.asoc.2011.10.014 bulletin of the polish academy of sciences, 51(3), 327-338.
  16. [16] http://users.pja.edu.pl/~wkos/11-KOSIN.PDF
  17. [17] Massanet, S., Riera, J. V., Torrens, J., & Herrera-Viedma, E. (2014). A new lin-guistic computational model based on discrete fuzzy numbers for computing with words. Information sciences, 258, 277-290. https://doi.org/10.1016/j.ins.2013.06.055
  18. [18] Nayagam, V. L. G., Jeevaraj, S., & Sivaraman, G. (2016). Complete ranking of intuitionistic fuzzy numbers. Fuzzy Information and engineering, 8(2), 237-254. https://doi.org/10.1016/j.fiae.2016.06.007
  19. [19] Nehi, H. M. (2010). A new ranking method for intuitionistic fuzzy numbers. International journal of fuzzy systems, 12(1).https://www.researchgate.net/publica-tion/279900044_A_New_Ranking_Method_for_Intuitionistic_Fuzzy_Numbers
  20. [20] Panda, A., & Pal, M. (2015). A study on pentagonal fuzzy number and its corre-sponding matrices. Pacific science review B: humanities and social sciences, 1(3), 131-139. https://doi.org/10.1016/j.psrb.2016.08.001
  21. [21] Qiu, D., & Zhang, W. (2013). Symmetric fuzzy numbers and additive equivalence of fuzzy numbers. Soft computing, 17(8), 1471-1477.https://doi.org/10.1007/s00500-013-1000-3
  22. [22] Rouhparvar, H., & Panahi, A. (2015). A new definition for defuzzification of gen-eralized fuzzy numbers and its application. Applied Soft Computing, 30, 577-584. https://doi.org/10.1016/j.asoc.2015.01.053
  23. [23] Seresht, N. G., & Fayek, A. R. (2019). Computational method for fuzzy arith-metiac operations on triangular fuzzy numbers by extension principle. International journal of approximate reasoning, 106, 172-193. https://doi.org/10.1016/j.ijar.2019.01.005
  24. [24] Smarandache, F. (1999). A unifying field in logics neutrosophy: neutrosophic probability. Set and Logic.
  25. [25] Wang, Y. M., Yang, J. B., Xu, D. L., & Chin, K. S. (2006). On the centroids of fuzzy numbers. Fuzzy sets and systems, 157(7), 919-926. https://doi.org/10.1016/j.fss.2005.11.006
  26. [26] Xu, Z., Shang, S., Qian, W., & Shu, W. (2010). A method for fuzzy risk analysis based on the new similarity of trapezoidal fuzzy numbers. Expert Systems with Applications, 37(3), 1920-1927. https://doi.org/10.1016/j.eswa.2009.07.015
  27. [27] Yager, R. R. (1986). On the theory of bags. International journal of general system, 13(1), 23-37. https://doi.org/10.1080/03081078608934952
  28. [28] Zad´eh, L.A. (1965) ‘Fuzzy sets’, Information and control, 8(3), 338–353.
  29. [29] Zadeh, L. A. (1975). The concept of a linguistic variable and its application to ap-proximate reasoning—I. Information sciences, 8(3), 199-249. https://doi.org/10.1016/0020-0255(75)90036-5