Sunayana Saikia
Department of Mathematics, Cotton University, Guwahati, Assam, India
In most real-world applications, single-valued neutrosophic fuzzy numbers are employed to make judgments in a fuzzy environment. As a result, single-valued neutrosophic fuzzy number ranking is crucial. This paper proposes a novel approach for ranking single-valued neutrosophic fuzzy numbers. The approach was created by combining the α, β, and γ cuts of neutrosophic fuzzy numbers in a convex way. In addition, the proposed approach has been used to prove a number of theorems. This paper also offers a number of numerical examples.
Keywords: Single valued neutrosophic fuzzy numbers, α-cut set, β-cut set, γ-cut set, ranking of neutrosophic fuzzy numbers, convex combination
Real-life phenomena cannot be classified by real numbers due to the existence of linguistic concepts. Fuzzy numbers can be used to rate such issues because fuzzy numbers may deal with verbal words that are ambiguous or vague. As a result, numerous ways for ranking fuzzy numbers have been presented at various times. In addition, the ordering of fuzzy numbers has been thoroughly examined. In the literature, there are several research articles that describe various strategies for sorting fuzzy numbers. In a nutshell, no universally accepted method for sorting fuzzy numbers exists. Existing methods have been demonstrated to give outcomes that are counter-intuitive at times. Furthermore, some of these strategies are non-discrimi-natory and counter intuitive. In 1965, Zadeh [33] “was the first to introduce the notion of fuzzy set theory”. Many academics have worked hard to come up with new and better techniques to rank the unclear numbers. Jain [18, 19] introduced the first unorthodox efforts to the condition of “ordering fuzzy quantities using maximizing set” in 1976. After that, “some ranking methods of fuzzy numbers examined and compared” by Bortolan and Degani [4] and Burnelli and Mezei [5]. In addition, Wang and Kerre [30, 31], Dubois and Prade [15], Chen [6], Choobineh and Li [9], Yu and Dat [32], Chen and Sanguansat [7], Chen et al. [8], Shureshjani and Darehmiraki [25], Chutia [10], Chutia and Chutia [11], Chutia and Gogoi [12], and so forth proposed some ranking methods and some properties of the ranking methods. In 1998, Smarandache [26–28] did the first pioneering work on neutrosophic fuzzy numbers. Ulucay et al. [29], Karaaslan [23], Deli [13], Jana and Pal [21], Jana et al. [20, 22], and others utilized the neutrosophic number for a variety of decision-making situations. Several studies have been conducted on the ordering of single-valued neutrosophic numbers. Peng et al. [24] established a single-valued neutrosophic number ranking system that was subsequently used to “multi-criteria decision making issues”. To rank the single-valued neutrosophic numbers, Deli and Subas [14], Biswas et al. [3], and Aal et al. [1] “presented values and ambiguities of truth membership, indeterminacy-membership, and falsity-membership functions”. Garai et al. [34] “devised a single-valued neutrosophic number ranking system based on possible mean, variance, and standard deviation, which they subsequently applied to multi-attribute group decision-making difficulties”. Garai et al. [16] “proposed a weighted possibility means-based ranking method for single-valued neutrosophic numbers”. Bhaumik et al. [2] “proposed a new ranking method for single-valued neutrosophic numbers established on the (α, β, γ)-cut set, which was then used to the solution of bi-matrix games”.
Because of their linguistic terms, ranking fuzzy numbers plays a distinguishing part in the decision-making process. In some decision-making scenarios, rating single-valued neutrosophic numbers is crucial. Single-valued neutrosophic numbers can be ranked using a variety of approaches. Furthermore, existing single-valued neutrosophic fuzzy number ranking algorithms are incorrect in several instances. As a result, new ways are required to overcome present methods’ limits and downsides. This study makes an effort at doing so. The suggested technique is based on a convex combination of single-valued neutrosophic numbers from the α, β, and γ cut sets.
In Section 24.2, some fundamental definitions and arithmetic of single valued neutrosophic fuzzy numbers are covered. The convex combination of α-cut, β-cut, and γ-cut sets of a single valued neutrosophic fuzzy number are proposed in Section 24.3 as a novel ranking technique for single valued neutrosophic fuzzy numbers. With the aid of the suggested ranking mechanism, we establish certain theorems in Section 24.4. Some numerical examples of the suggested technique are presented in Section 24.5. The proposed method’s conclusion and future studies has been presented in Section 24.6.
This section looked at several definitions and representations for the idea of single-valued neutrosophic fuzzy numbers.
Definition 24.2.1 ([11]). “Let A represent a set and X represent the universal set, with A ∈ X. The set of ordered pairs A = {(x, µA(x)): x ∈ X, µA,: X → [0, 1]} is then referred to as a fuzzy set and µA is the membership function”.
Definition 24.2.2 ([11]). “Let a fuzzy number A = (x, µA(x)) = (a, b, c, d; wA) and if the membership function µA(x) is represented as,
then the fuzzy number A is called a generalized trapezoidal fuzzy number (GTrFN)”.
Definition 24.2.3 ([11]). “An intuitionistic fuzzy set (IFS) A on ℝ is defined as A = ⟨x, µA(x), νA(x)⟩: x ∈ ℝ), where the function µA: ℝ → [0, 1] is known as the degree of membership and νA: ℝ → [0, 1] is known as the degree of non-membership of the element x ∈ ℝ, such that 0 ≤ µA(x) + νA(x) ≤ 1”.
Definition 24.2.4 ([11]). “Consider an intuitionistic fuzzy number (IFN) A = ⟨(a, b, c, d; wA), (p, q, r, s; uA)⟩, where c ≤ a, q ≤ b ≤ c ≤ r, d ≤ s are the parameters. The membership function and the non-membership function are defined as:
respectively, where wA is the maximum degree of membership and uA is the minimum degree of non-membership such that 0 ≤ wA ≤ 1, 0 ≤ uA ≤ 1 and 0 ≤ wA + uA ≤ 1”.
Definition 24.2.5 ([17]). “Assume that X is the universal set and that A is a set in X. The ordered pair set A = {⟨x, µA(x), νA(x), ρA(x)⟩: x ∈ X} is a single-valued neutrosophic set (SVNS), with µA(x) being the truth membership degree, defined as µA: X → [0, 1], νA(x) being the indeterminacy membership degree, defined as νA: X → [0, 1] and ρA(x) being the falsity membership degree is defined as ρA: X → [0, 1], of the element x to the set A, with the constraint 0 ≤ µA(x) + νA(x) + ρA(x) ≤ 3, ∀x ∈ X”.
Definition 24.2.6 ([17]). “A SVNS A = {⟨x, µA(x), νA(x), ρA(x)⟩∀x ∈ X} is said to be neutrosophic-normal if there exist at least three points a, b, c ∈ X where µA: a → {1}, νA: b → {1} and ρA: c → {1}”.
Definition 24.2.7 ([17]). “An (α, β, γ)-cut set of a SVNS A is a crisp subset over the set of real numbers ℝ which are defined as, A(α, β, γ) = {x,: µA(x) ≥ α, νA(x) ≤ β, ρA(x) ≤ γ}, with 0 ≤ α ≤ 1, 0 ≤ β ≤ 1 and 0 ≤ α + β + γ ≤ 3”.
Definition 24.2.8 ([17]). “A SVNS, A = {⟨x, µA(x), νA(x), ρA(x)⟩∀x ∈ X} is called neutrosophic-convex if:
Definition 24.2.9 ([17]). “A single-valued nutrosophic set {A = ⟨x, µA(x), νA(x), ρA(x)⟩∀x ∈ X} is said to be a single-valued neutrosophic number (SVNN) if,
Definition 24.2.10 ([17]). “A SVNS, A = {⟨x, µA(x), νA(x), ρA(x)⟩∀x ∈ X},= ⟨(a1, b1, c1), (i1, j1, k1), (p1, q1, r1)⟩ in the set of real numbers is said to be single-valued triangular neutrosophic number (SVTNN), such that p1 ≤ i1 ≤ a1 ≤ q1 ≤ j1 ≤ b1 ≤ c1 ≤ j1 ≤ r1 are the parameters. Then, the truth membership degree, the indeterminacy membership degree, and thee falsity membership degree of A can be defined as:”
Definition 24.2.11 ([2]). Cut Sets of SVTNN
“Let, a SVTNN A = {⟨µA(x), νA(x), ρA(x)⟩: x ∈ X} = ⟨(a1, b1, c1), (i1, j1, k1), (p1, q1, r1)⟩, where a1, b1, c1, i1, j1, k1, p1, q1, r1 ∈ R. Then, (α, β, γ)-cut sets of A are defined as:
Definition 24.2.12 ([2]). Defined the Quantity Value
“Let A = {⟨µA(x), νA(x), ρA(x)⟩: x ∈ X} be a SVNN. Then,
where α ∈ [0, 1] and the weighted function f (α) = α satisfies the condition f (0) = 0 and f (1) = 1, such that .
where β ∈ [0, 1] and the weighted function g(β) = 1 − β satisfy the condition g(0) = 1 and g(1) = 0, such that .
where γ ∈ [0, 1] and the weighted function h(γ) = 1 − γ satisfy the condition h(0) = 1 and h(1) = 0, such that .
Definition 24.2.13 ([2]). Arithmetic
Let two SVTNNs as A = {⟨µA(x), νA(x), ρA(x)⟩: x ∈ X} = ⟨(a1, b1, c1), (i1, j1, k1), (p1, q1, r1)⟩ ∈ ℝ and B = {⟨µB(x), νB(x), ρB(x)⟩: x ∈ X} = ⟨(a2, b2, c2), (i2, j2, k2), (p2, q2, r2)⟩ ∈ ℝ, where µA and µB are truth memberships, νA and νB are indeterminacy memberships, and ρA and ρB are falsity memberships of A and B, respectively. Then,
Here, a new ranking method for single-valued neutrosophic fuzzy numbers (SVNFNs) has been discussed. Consider a fuzzy number A ∈ X to be a SVTNN, such as A = ⟨x, µA(x), νA(x), ρA(x)⟩ = ⟨[Lα(A), Rα(A)], [Lβ(A), Rβ(A)], [Lγ(A), Rγ(A)]⟩, 0 ≤ α, β, γ ≤ 1, “where [Lα(A), Rα(A)] is α- cut set, [Lβ(A), Rβ(A)] is β- cut set and [Lγ(A), Rγ(A)] is γ-cut set for truth membership function, indeterminacy membership function, and falsity membership function” of the SVTNN A, respectively.
Then, the proposed ranking index at α, β, γ cuts is defined as
where λ1, λ2, λ3 ∈ [0, 1].
Thus, the ordering of the fuzzy numbers A, B ∈ X, based on the ranking index is defined as:
where (0 ≤ λ1, λ2, λ3 ≤ 1).
Some crucial and fundamental theorems are discussed here.
Theorem 24.4.1. Let be linear.
Proof. Let, A = ⟨(a1, b1, c1), (i1, j1, k1), (p1, q1, r1)⟩ and B = ⟨(a2, b2, c2), (i2, j2, k2), (p2, q2, r2)⟩ be two SVTNNs and κ be any real number. Then, it has to prove that . Then,
Theorem 24.4.2. If A, B, and C are three arbitrary SVTNNs and A ≲ B and B ≲ C are true, then A ≲ C is true.
Proof. Given that A ≲ B, then (A) ≤ (B) for 0 ≤ λ1, λ2, λ3 ≤ 1. Also, given that B ≲ C, then (B) ≤ (C), for 0 ≤ λ1, λ2, λ3 ≤ 1. This implies that (A) ≤ (C), for 0 ≤ λ1, λ2, λ3 ≤ 1. Hence, the result follows that A ≲ C.
In this section, some numerical examples are discussed to assess how well the proposed method performs.
Example 24.5.1. Consider two SVTNNs A = ⟨(0.50, 0.65, 0.80), (0.10, 0.15, 0.30), (0.10, 0.20, 0.30)⟩, B = ⟨(0.10, 0.20, 0.30), (0.20, 0.30, 0.40), (0.40, 0.50, 0.70)⟩ and three weighted functions be P = (λ1 = 0.3, λ2 = 0.4, λ3 = 0.5), Q = (λ1 = 0.5, λ2 = 0.5, λ3 = 0.5) and S = (λ1 = 1.0, λ2 = 1.0, λ3 = 0.9).
Then, the comparison is between two SVTNNs, A and B, as shown in Table 24.1.
Example 24.5.2. Consider two SVTNNs A = ⟨(0.30, 0.45, 0.50), (0.10, 0.20, 0.40), (0.10, 0.20, 0.30)⟩, B = ⟨(0.20, 0.30, 0.35), (0.10, 0.10, 0.10), (0.60, 0.70, 0.80)⟩ and three weighted functions P = (λ1 = 1.0, λ2 = 1.0, λ3 = 1.0), Q = (λ1 = 0.5, λ2 = 0.5, λ3 = 0.5) and S = (λ1 = 0.1, λ2 = 0.1, λ3 = 0.1).
Then, the comparison is between the two SVTNNs, A and B, as shown in Table 24.2.
Example 24.5.3. Consider two SVTNNs A = ⟨(0.6, 0.7, 0.9), (0.1, 0.3, 0.4), (0.1, 0.4, 0.6)⟩, B = ⟨(0.1, 0.2, 0.3), (0.4, 0.5, 0.6), (0.7, 0.8, 0.9)⟩ and three weighted functions P = (λ1 = 0.3, λ2 = 0.4, λ3 = 0.5), Q = (λ1 = 0.3, λ2 = 0.6, λ3 = 0.5) and S = (λ1 = 0.6, λ2 = 0.4, λ3 = 0.7). Then, the comparison is between the two SVTNNs, A and B, as shown in Table 24.3.
Table 24.1 Ranking of SVNNs in Example 24.5.1.
λ1,λ2,λ3, | (A) | (B) | Results |
---|---|---|---|
0.3,0.4,0.5 | 0.5217 | 0.5384 | A ≺ B |
0.5,0.5,0.5 | 0.5083 | 0.5083 | A ~ B |
1.0,1.0,0.9 | 0.4534 | 0.45502 | A ≻ B |
Table 24.2 Ranking of SVNNs in Example 24.5.2.
λ1,λ2,λ3, | (A) | (B) | Results |
---|---|---|---|
1.0,1.0,1.0 | 0.0669 | 0.5168 | A ≺ B |
0.5,0.5,0.5 | 0.4251 | 0.5460 | A ≺ B |
0.1,04,0.1 | 0.4719 | 0.5693 | A ≺ B |
Table 24.3 Ranking of SVNNs in Example 24.5.3.
λ1,λ2,λ3, | (A) | (B) | Results |
---|---|---|---|
0.3,0.4,0.5 | 0.6584 | 0.8101 | A ≺ B |
0.3,0.6,0.5 | 0.7550 | 0.7500 | A ≻ B |
0.6,0.4,0.7 | 0.7000 | 0.7401 | A ≺ B |
The majority of real-life decision-making is based on fuzzy numbers due to their language connotations. Ordering fuzzy numbers is much more important in order to solve such difficulties. There are several documented ranking systems for single-valued fuzzy numbers. This paper developed a novel ranking approach for single-valued fuzzy numbers based on convex combinations of the fuzzy numbers’ α, β, and γ cut sets. A number of theorems have also been proved using the proposed method. Furthermore, numerical examples have been provided to demonstrate the suggested strategy’s superior performance. This ranking method may be used to look at the rankings of intuitionistic fuzzy numbers, type-2 fuzzy numbers, hesitant fuzzy numbers, and other fuzzy numbers. Further, this proposed method can be applied to real-life problems like medicine, business, economics, computer sciences, etc.