Introduction
Due to the increased emergencies, health care systems etc., decision-making and information management have become very important. However, to handle complex, imprecise and ambiguous information [
1,
2], it isn’t easy to use classical tools. Therefore, to overcome this situation, Zadeh introduced the theory of fuzzy set (FS) [
3]. FS is characterized by membership function (MF) in the closed interval [0, 1]. For transportation problems, an approach namely, P. Senthil Kumar (PSK) method was developed in order to obtain an optimal solution in terms of triangular fuzzy numbers [
4,
5].
The theory of FS was further extended by Al-shami et al. [
6] and introduced the notion of (a, b)-fuzzy soft set. Further, some aggregation operators are developed in order to aggregate the (a, b)-fuzzy soft set. Ibrahim et al. [
7] initiated the notion of (3,2)-fuzzy sets and discussed their relationship with other kinds of fuzzy sets.
Further, Atanassov extended the notion of FS and initiated the concept of an intuitionistic fuzzy set (IFS) [
8]. The IFS is characterized by MF and nonmembership function (NMF) belonging to the closed interval [0, 1] such that the
\(MF + NMF \le 1\). For aggregating intuitionistic fuzzy numbers, Xu and Yager [
9] presented a geometric aggregation operator. In addition, related to IFS, some authors proposed several other tools to handle vague and imprecise information whereby two or more sources of vagueness appear simultaneously [
10‐
14]. The notion of SR-Fuzzy Sets was first introduced by Al-Shami et al. [
15] as a generalization of FS. Further, the authors developed weighted aggregated operators for the aggregation of SR-Fuzzy sets and applied the concept to MCDM problem. Moreover, Al-shami and Mhemdi [
16] introduced the concept of (m, n)-fuzzy sets and proposed an MCDM approach with the (m, n)-fuzzy set setting Al-Shami [
17] further initiated the concept of (2,1)-Fuzzy sets and developed some weighted aggregated operators.
Kumar [
18] developed an algorithm for solving real-life optimization problems based on fuzzy and IFS. Kumar [
19] developed a software-based approach for solving the intuitionistic fuzzy solid assignment problem (IFSAP). A numerical example was also presented for the application of the proposed method. In order to deal with uncertainty and hesitation in transportation problems an intuitionistic fuzzy zero point method was developed by Kumar [
20] to find the optimal solution to the type-2 intuitionistic fuzzy transportation problem (IFTP). Kumar [
21] proposed a computationally simple and efficient method for solving real-life mixed intuitionistic fuzzy 3D assignment problems. Further, Kumar [
22] further formulated a transportation problem in which costs are triangular intuitionistic fuzzy numbers, and supplies and demands are crisp numbers.
Atanassov and Gargov [
23] introduced an interval-valued intuitionistic fuzzy set (IIFS) characterized by a membership function and a nonmembership function with interval values. Xu [
24] presented an averaging operator (AO) to deal with multi-criteria decision-making (MCDM) problems in the IIFS environment. Garg used Einstein’s t-norm techniques to create numerous interactive AO for interval-valued intuitionistic fuzzy numbers (IIFNs) [
25,
26]. The ordered weighted aggregate (OWA) operator was initiated by Yager [
27], adding weights to every input depending on their ranking positions. Garg et al. [
28] used Hamacher TN operators to provide different AOs under IIFS. For IIFN aggregation, Wang and Liu [
29] suggested an Einstein hybrid weighted geometric aggregation operator. Liu [
30] presented IIFN AOs based on Hamacher norms. Chen et al. [
31] introduced several new AOs for solving the MCDM problem in the IIFS context based on transformation approaches. Triangular norms play an important role and are the basis of many AOs addressed in many fuzzy frameworks. Menger [
32] pioneered the use of triangular norms in statistical data. Deschrijver et al. [
33] studied tnorm (TN) and tconorm (TCN) in the context of IFSs. Several triangular norms were established to collect knowledge in various mathematical frameworks. Lukasiewicz TN and TCN [
34], product TN and probabilistic sum TCN [
35], Archimedean TN and TCN [
36], Drastic TN and TCN [
37], Einstien TN and TCN [
38], and Dombi TN and TCN [
39] are some of the TN and TCN that have been proposed. Nevertheless, in complex data sets, ambiguity and vagueness within the data occur concurrently with changes in the data’s phase (periodicity). Complex data sets include vast amounts of data generated by medical research and government databases for biometric and face identification, sound and photographs, all of which may contain large amounts of incomplete, confusing, and inaccurate information. To overcome these concerns, Ramot et al. [
40] proposed the concept of a complex fuzzy set (CFS), which expands the membership function’s range from the real number subset to the unit disc. Ramot et al. [
40] investigated several characteristics of CFSs, such as complements, unions, and intersections, using many illustrative instances. Because the CFS model does not tell anything about the disagreement of any element in any setting, Alkouri and Salleh [
41] extended the concept of CFS to complex intuitionistic fuzzy set (CIFS) by including the degree of non-membership and defining their primary operations, such as union, intersection, complement, and so on. Alkouri and Salleh [
42] also defined complex intuitionistic fuzzy connection, structure, and projections, and even a similarity measure among two CIFSs. Kumar and Bajaj [
43] gave many distance and entropy metrics for complex intuitionistic fuzzy soft sets. Rani and Garg [
44] proposed a set of distance measures for use in decision-making in a CIFS context. Rani and Garg [
45] developed power AOs to integrate the various CIFS and handle the MCDM issue. Garg and Rani [
46] proposed several general aggregating operators for different CIFS. In 1982, Aczel and Alsina [
47] developed a novel family of TN and TCN for the condition
\(0\le p\le \infty \) in 1982, dubbed Aczel–Alsina TN and TCN. When the value of p increases, the AATN and AATCN are strictly growing and continuous. Many scholars have utilised the ideas of AATN and AATCN to discover the superiority of modifying active parameters in many sectors. Different triangular norms of parametric TN, product TN, Dombi product TN, AATN, Frank product TN, and Schweizer and Sklar TN were studied by Babu and Ahmed [
48].
Senapati and Chen [
49] first defined Aczel-Alsina aggregation operators for IFS. He also defined [
50] the same operators for IIFS. In [
51], Mehmood et al. developed some complex intuitionistic fuzzy aggregation operators based on Aczel-Alsina t-norm and t-conorm. However, complex intuitionistic fuzzy set (CIFS) deal only with single-valued membership degrees but cannot deal with interval values. Therefore, to give experts more freedom, Rani and Garg [
52] introduced the concept of complex interval-valued intuitionistic fuzzy set (CIIFS). CIIFS is characterized by a complex-valued membership degree and a complex-valued non-membership degree range from the real number subset to the unit disc. Motivated by the CIIFS and Aczel-Alsina t-norm and t-conorm in this paper, some new Aczel-Alsina t-norm and t-conorm based operators namely the complex interval-valued intuitionistic fuzzy Aczel-Alsina t-norm and t-conorm (CIIFAA) operators are developed. Then some properties of the proposed operators are discussed, and the concept is applied to multi-criteria decision-making problems. The following are the key contributions of this paper:
(1)
To define operational laws based on Aczel-Alsina t-norm and t-conorm for complex interval-valued intuitionistic fuzzy numbers (CIIFNs).
(2)
To develop Aczel-Alsina t-norm and t-conorm based aggregation operators for CIIFNs.
(3)
To propose an MCDM method based on the developed operators.
(4)
A case study, a numerical example of COVID-19 healthcare facilities, is given to show how the proposed operators can be used.
The rest of the text reads as follows: Section “
Preliminaries” presents some basic definitions that are preliminary results. Section “
Operational laws for CIIFNs based on A-A t-norm and A-A t-conorm” deals with some operational laws for CIIFNs based on Aczel-Alsina t-norm and t-conorm. Its order and hybrid form are proposed using the proposed operational laws in Section “
CIIF Aczel-Alsina average aggregation operators”, complex interval-valued intuitionistic fuzzy Aczel-Alsina t-norm and t-conform-based weighted average (CIIFAAWA) operator. Some properties of the proposed operators are also discussed in detail. Section “
CIIF Aczel-Alsina geometric aggregation operators” investigates a complex interval-valued intuitionistic fuzzy Aczel-Alsina t-norm and t-conorm-based weighted geometric (CIIFAAWG) operator, its order, hybrid form, and the properties of the proposed operators. Section “
Model for MADM using CIF data” presents an MCDM approach based on the developed operators in a CIIFS environment. Section “
Health care facility for COVID-19 in hospitals” is a numerical example regarding COVID-19 facilities in public sector hospitals for demonstrating and applying the proposed method. Section “
Conclusions” presents a concluding remark.
CIIF Aczel-Alsina average aggregation operators
We develop the succeeding theorem that follows the Aczel-Alsina operations on CIIFNs.
We easily prove the subsequent properties by applying the CIIFAAWA operator.
Now, we introduce CIF Aczel-Alsins ordered weighted averaging (CIFAAOWA) operator.
In Definition 2, we comprehend that the CIIFAAWA operator weights the CIIFNs, and in Definition 3, we recognize that the CIIFAAOWA operator weights only the ordered position of the CIIFNs. As such, the above-mentioned two average operators indicate the weights in two ways. But the weight can be shown by one operator without being directed in two ways by two operations. To overcome this difficulty, we put forward CIIF Aczel-Alsina hybrid averaging (CIIFAAHA) operator, which weighs both the given CIIFN and its ordered position.
Based on Aczel-Alsina operations with CIIFNs information, we can deduce the subsequent theorem.
Just like the above Theorem, we can easily obtain the following theorem.
CIIF Aczel-Alsina geometric aggregation operators
We develop the succeeding theorem that follows the Aczel-Alsina operations on CIIFNs.
We easily prove the subsequent properties by using the CIFAAWG operator.
We introduce the CIF Aczel-Alsins ordered weighted geometric (CIIFAAWG) operator.
The CIIFAAOW operator may show the subsequent properties without any problem.
In the above definition, we comprehend that the CIIFAAWG operator weights solely the CIIFNs. In this definition, the CIIFAAOWG operator only considers the CIIFNs’ ordered position. As such, the two above geometric operators indicate the weights in two ways. But the weight can be shown by one operator without being directed in two directions by two operations. We suggest using the CIIF Aczel-Alsina hybrid geometric (CIIFAAHG) operator to solve this problem. This operator takes into account both the given CIIFN and its ordered position.
We can deduce the subsequent theorem based on Aczel-Alsina operations with CIIFNs information.
From the above theorem, we can easily obtain the following theorem:
Model for MADM using CIF data
This section will utilize the recommended operators for a MADM issue in the CIF environment. MADM techniques make decisions by analyzing trade among alternative exhibitions over multiple attributes [
30]. Basic data required in a MADM technique incorporates attribute values or performance measures (individual assessments on alternatives against each attribute), attribute weights (which take into account how important each attribute is to the whole decision issue), and a way to put all of this information together into a total value or assessment for each option. Simply put, a MADM problem is finding a good compromise solution out of all the possible options after weighing them on multiple factors. Here, we might also advise a MADM process controlling CIF Aczel Alsina aggregation operators where characteristic values are inside the shape of CIFEs, and characteristic weights are inside the shape of actual numbers. Assume that
\(\Gamma _{CIIF}=\{ \Gamma _{CIIF_{(1)}},\Gamma _{CIIF_{(2)}}, \ldots ,\Gamma _{CIIF_{(m)}}\}\) are the discretely preparations of alternatives,
\(\chi =\{ \chi _{1},\chi _{2}, \ldots ,\chi _{n}\}\)are the discretely preparations of attributes and ’
\(\varphi =(\varphi _{1},\varphi _{2}, \ldots ,\varphi _{n})\) are the weight vector of the characteristic
\(\chi _{\lambda }(\lambda =1,2, \ldots ,n)\) in any such manner as ’
\(\varphi _{\lambda }\in [0,1]\) and
\(\sum _{\lambda =1}^{n}=1\). Suppose that R
\(=(CIIF_{s(\lambda )})_{m\times n}\) \(=\) \((\Gamma _{CIIF_{s(\lambda )}},\Lambda _{CIIF_{s(\lambda )}})_{m\times n}\) is the CIF selection matrix, as seemed in Table
1, where
\(\Gamma _{\acute{C}_{CIIF_{s(\lambda )}}}\) in constitute club diploma with the aid of using which opportunity
\(\Gamma _{CIIF_{s}}\) fulfils the characteristic
\(\chi _{\lambda }\), and
\(\Lambda _{CIIF_{\acute{C} _{CIIF_{s(\lambda )}}}}\) constitute non-club diploma with the aid of using which opportunity
\(\mu _{\lambda }\) does now no longer fulfil the characteristic
\(\chi _{\lambda }\), where
\(\Gamma _{CIIF_{\acute{C}_{CIIF_{s(\lambda )}}}}\subset [0,1]\), and
\(\Lambda _{CIIF_{\acute{C}_{CIIF_{s(\lambda )}}}}\subset [0,1]\) permitting
\(0\le \Gamma _{CIIF_{\acute{C}_{CIIF_{s(\lambda )}}}}+\Lambda _{CIIF_{\acute{C}_{CIIF_{s(\lambda )}}}}\le 1\),
\((s=1,2, \ldots ,m)\) and
\((\lambda =1,2, \ldots ,n).\)Table 1
Decision making matrix \(R=(CIIF_{s(\lambda )})_{m\times n}\)
\((\Gamma _{CIIF_{11},\Lambda _{CIIF_{11}}})\) | \((\Gamma _{CIIF_{11},\Lambda _{CIIF_{11}}})\) | ... | \((\Gamma _{CIIF_{11},\Lambda _{CIIF_{11}}})\) |
\((\Gamma _{CIIF_{11},\Lambda _{CIIF_{11}}})\) | \((\Gamma _{CIIF_{11},\Lambda _{CIIF_{11}}})\) | ... | \((\Gamma _{CIIF_{11},\Lambda _{CIIF_{11}}})\) |
\(\vdots \) | \(\vdots \) | \(\ddots \) | \(\vdots \) |
\((\Gamma _{CIIF_{11},\Lambda _{CIIF_{11}}})\) | \((\Gamma _{CIIF_{11},\Lambda _{CIIF_{11}}})\) | ... | \((\Gamma _{CIIF_{11},\Lambda _{CIIF_{11}}})\) |
The following algorithm constructs a method for searching for the best opportunity(s) in light of the CIIFAAWA operators for MADM difficulty, which includes the following steps:
Step I: For a MADM difficulty with CIIFEs, we formulate the CIIF selection matrix \(R=(\acute{C}_{CIF_{s(\lambda )}})_{m\times n}\) where the factors \(\acute{C}_{CIF_{s(\lambda )}}\) \((s=1,2, \ldots ,m,\lambda =1,2, \ldots ,n)\) are the value determinations of the opportunity \(\mu _{m}\) concerning the criterion \( \chi _{n}\).
Step II: In the occasion that there are numerous varieties of criteria, in particular benefit (B) and cost (C), at that factor, we convert the CIIF selection matrix \(R=(\acute{C}_{s\lambda })_{m\times n}\) into the normalized one\(R^{\sim }\) \(=(\mu _{_{s\lambda }})_{m\times n}\) via way of means of making use of the accompanying equation \(\Gamma _{CIF_{s\lambda }}=\left\{ \begin{array}{c} \acute{C}_{CIF_{s(\lambda )}},\lambda \in B \\ \acute{C}_{CIF_{s(\lambda )}}^{c},\lambda \in C_{CIF} \end{array} \right\} \), where \(\acute{C}_{CIF_{s(\lambda )}}^{c}\) is the compliment of \(\acute{C} _{CIF_{s(\lambda )}}\).
Step III: On the basis of the decision matrix
\(R^{\sim }\), as derived from Step 2, the total accumulated value of the alternative
\(\Gamma _{CIF_{s}}(s=1,2,\ldots ,m)\) under the different criteria
\(\chi _{\lambda }\)is acquired by utilizing CIIFAAWA operator and hence get the general decision values
\(\Gamma _{CIF_{s}}(s=1,2,\ldots ,m)\) for each alternative
\(\mu _{s}\), i.e.,
Step IV: We decide the rating values
\(S^{\sim }(\Gamma _{CIF_{s}})(s=1,2,\ldots ,m)\) at the basis of the overall CIF records
\(\Gamma _{CIF_{s}}(s=1,2,\ldots ,m)\) with the reason of ranking all of the opportunity
\(\Gamma _{CIF_{s}}(s=1,2,\ldots , m) \) to choose wonderful desire
\(\Gamma _{CIF_{s}}\). If you discover equal rating capabilities
\(S^{\sim }(\Gamma _{CIF_{s}})\) and
\(S^{\sim }(\Gamma _{CI_{IF}})\), at that factor we retain to parent accuracy tiers of
\(L^{\sim }(\Gamma _{CIF_{s}})\) and
\(L^{\sim }(\Gamma _{CI_{IF}})\) in view of standard CIF records of
\(\Gamma _{CIF_{s}}\) and
\(\Gamma _{CI_{IF}}\), and we rank the selection
\(\Gamma _{CIF_{s}}\) consistent with the accuracy tiers of
\(L^{\sim }(\Gamma _{CIF_{s}})\) and
\(L^{\sim }(\Gamma _{CI_{IF}})\).
Step V: We rank all of the opportunity \(\Gamma _{CIF_{s}}(s=1,2,\ldots ,m)\) in accordance with the descending price of the rating values, choose the most suited opportunity.
Step VI: End.
Health care facility for COVID-19 in hospitals
COVID-19 is the name of the new coronavirus strain driving the current epidemic. On January 7, 2020, Chinese authorities identified this virus strain [
53]. It began to spread worldwide from country to country [
54], and the number of cases grew by the day [
55]. Every year, the global health budget is $7.5 trillion [
53]. This epidemic has served as a harsh wake-up call for Pakistan’s inadequate health infrastructure, which has been under unsustainable strain throughout this time [
56]. The COVID-19 pandemic for 2019–2020 was expected to enter Pakistan in February 2020, with approximately 255,769 cases and 5386 deaths (World Health Organization 2020b) as of July 15, 2020. In a developing country like Pakistan, disease outbreaks are a big problem for the health care system. There aren’t enough basic medical facilities or health policies, the government isn’t doing a good job, and people aren’t interested in broad protective measures [
57].
This paper presents a numerical example regarding healthcare facilities in Khyber Pakhtunkhwa, Pakistan. After initial scrutiny, four hospitals in Khyber Pakhtunkhwa are taken as alternatives
\(A_{i}=(i=1,2,3,4,5).\) The four hospitals are evaluated based on the following four criteria:
1.
RT-PCR Facilities (RT-PCR),
2.
Personal protective equipment (PPE),
3.
Shortage of Isolation Ward (SIW),
4.
Hight Qualified Staff (HQS).
A group of three experts is nominated in order to evaluate the four hospitals based on four criteria. The experts present their evaluation values in the form of CIIFNs, indicated in Tables
2,
3,
4.
Table 2
Complex interval-valued intuitionistic fuzzy decision matrix by \(E_{1}\)
\(\hbox {A}_{{1}}\) | \( \left( \begin{array}{c}\,\left[ 0.3,0.4\right] e^{i[2\pi \left( 0.1\right) , 2\pi \left( 0.3\right) ]}, \\ \,\left[ 0.1,0.3\right] e^{i[2\pi \left( 0.3\right) , 2\pi \left( 0.4\right) ]}\end{array}\right) \) | | \( \left( \begin{array}{c}\,\left[ 0.4,0.5\right] e^{i[2\pi \left( 0.2\right) , 2\pi \left( 0.4\right) ]}, \\ \,\left[ 0.3,0.5\right] e^{i[2\pi \left( 0.1\right) , 2\pi \left( 0.2\right) ]}\end{array}\right) \) | |
\(\hbox {A}_{{2}}\) | \( \left( \begin{array}{c}\,\left[ 0.1,0.3\right] e^{i[2\pi \left( 0.3\right) , 2\pi \left( 0.4\right) ]}, \\ \,\left[ 0.3,0.4\right] e^{i[2\pi \left( 0.1\right) , 2\pi \left( 0.3\right) ]}\end{array}\right) \) | \( \left( \begin{array}{c}\,\left[ 0.3,0.4\right] e^{i[2\pi \left( 0.2\right) , 2\pi \left( 0.5\right) ]}, \\ \,\left[ 0.1,0.4\right] e^{i[2\pi \left( 0.2\right) , 2\pi \left( 0.3\right) ]}\end{array}\right) \) | \( \left( \begin{array}{c}\,\left[ 0.2,0.4\right] e^{i[2\pi \left( 0.4\right) , 2\pi \left( 0.5\right) ]}, \\ \,\left[ 0.1,0.2\right] e^{i[2\pi \left( 0.3\right) , 2\pi \left( 0.5\right) ]}\end{array}\right) \) | \( \left( \begin{array}{c}\,\left[ 0.3,0.4\right] e^{i[2\pi \left( 0.1\right) , 2\pi \left( 0.3\right) ]}, \\ \,\left[ 0.1,0.2\right] e^{i[2\pi \left( 0.3\right) , 2\pi \left( 0.5\right) ]} \end{array} \right) \) |
\(\hbox {A}_{{3}}\) | \( \left( \begin{array}{c} \,\left[ 0.3,0.4\right] e^{i[2\pi \left( 0.1\right) , 2\pi \left( 0.3\right) ]}, \\ \,\left[ 0.1,0.2\right] e^{i[2\pi \left( 0.3\right) , 2\pi \left( 0.5\right) ]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.2,0.4\right] e^{i[2\pi \left( 0.4\right) , 2\pi \left( 0.5\right) ]}, \\ \,\left[ 0.1,0.2\right] e^{i[2\pi \left( 0.3\right) , 2\pi \left( 0.5\right) ]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.3,0.4\right] e^{i[2\pi \left( 0.2\right) , 2\pi \left( 0.5\right) ]}, \\ \,\left[ 0.1,0.4\right] e^{i[2\pi \left( 0.2\right) , 2\pi \left( 0.3\right) ]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.1,0.3\right] e^{i[2\pi \left( 0.3\right) , 2\pi \left( 0.4\right) ]}, \\ \,\left[ 0.3,0.4\right] e^{i[2\pi \left( 0.1\right) , 2\pi \left( 0.3\right) ]} \end{array} \right) \) |
\(\hbox {A}_{{4}}\) | \( \left( \begin{array}{c} \,\left[ 0.1,0.3\right] e^{i[2\pi \left( 0.3\right) , 2\pi \left( 0.4\right) ]}, \\ \,\left[ 0.3,0.5\right] e^{i[2\pi \left( 0.1\right) , 2\pi \left( 0.2\right) ]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.4,0.5\right] e^{i[2\pi \left( 0.2\right) , 2\pi \left( 0.4\right) ]}, \\ \,\left[ 0.3,0.5\right] e^{i[2\pi \left( 0.1\right) , 2\pi \left( 0.2\right) ]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.2,0.5\right] e^{i[2\pi \left( 0.3\right) , 2\pi \left( 0.4\right) ]}, \\ \,\left[ 0.2,0.3\right] e^{i[2\pi \left( 0.1\right) , 2\pi \left( 0.4\right) ]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.3,0.4\right] e^{i[2\pi \left( 0.1\right) , 2\pi \left( 0.3\right) ]}, \\ \,\left[ 0.1,0.3\right] e^{i[2\pi \left( 0.3\right) , 2\pi \left( 0.4\right) ]} \end{array} \right) \) |
Table 3
Complex interval-valued intuitionistic fuzzy decision matrix by \(E_{2}\)
\(\hbox {A}_{{1}}\) | \( \left( \begin{array}{c} \,\left[ 0.1,0.3\right] e^{i[2\pi (0.3),2\pi 0.4)]}], \\ \,\left[ 0.3,0.5\right] e^{i[2\pi (0.1),2\pi 0.2)]}] \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.4,0.5\right] e^{i[2\pi (0.2),2\pi (0.4)]} \\ \,\left[ 0.3,0.5\right] e^{i[2\pi (0.1),2\pi (0.2)]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.2,0.5\right] e^{i[2\pi (0.3),2\pi (0.4)]} \\ \,\left[ 0.2,0.3\right] e^{i[2\pi (0.1),2\pi (0.4)]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.2,0.5\right] e^{i[2\pi (0.3),2\pi (0.4)]} \\ \,\left[ 0.2,0.3\right] e^{i[2\pi (0.1),2\pi (0.4)]} \end{array} \right) \) |
\(\hbox {A}_{{2}}\) | \( \left( \begin{array}{c} \,\left[ 0.3,0.4\right] e^{i[2\pi (0.1),2\pi (0.3)]}, \\ \,\left[ 0.1,0.2\right] e^{i2\pi [(0.3),2\pi \left( 0.5\right) ]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.2,0.4\right] e^{i[2\pi (0.4),2\pi (0.5)]} \\ \,\left[ 0.1,0.2\right] e^{i[2\pi (0.3),2\pi (0.5)]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.3,0.4\right] e^{i[2\pi (0.2),2\pi (0.5)]} \\ \,\left[ 0.1,0.4\right] e^{i[2\pi (0.2),2\pi (0.3)]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.3,0.4\right] e^{i[2\pi (0.2),2\pi (0.5)]} \\ \,\left[ 0.1,0.4\right] e^{i[2\pi (0.2),2\pi (0.3)]} \end{array} \right) \) |
\(\hbox {A}_{{3}}\) | \( \left( \begin{array}{c} \,\left[ 0.1,0.3\right] e^{i[2\pi (0.3),2\pi \left( 0.4\right) ]} \\ \,\left[ 0.3,0.4\right] e^{i[2\pi (0.1),2\pi \left( 0.3\right) ]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.3,0.4\right] e^{i[2\pi (0.2),2\pi (0.5)]} \\ \,\left[ 0.1,0.4\right] e^{i[2\pi (0.2),2\pi (0.3)]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.2,0.4\right] e^{i[2\pi (0.4),2\pi (0.5)]} \\ \,\left[ 0.1,0.2\right] e^{i[2\pi (0.3),2\pi (0.5)]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.2,0.4\right] e^{i[2\pi (0.4),2\pi (0.5)]} \\ \,\left[ 0.1,0.2\right] e^{i[2\pi (0.3),2\pi (0.5)]} \end{array} \right) \) |
\(\hbox {A}_{{4}}\) | \( \left( \begin{array}{c} \,\left[ 0.3,0.4\right] e^{i2\pi \left( 0.1\right) , 2\pi (0.3)]}\\ \,\left[ 0.1,0.3\right] e^{i2\pi \left( 0.3\right) , 2\pi (0.4)]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.2,0.5\right] e^{i[2\pi (0.3),2\pi (0.4)]} \\ \,\left[ 0.2,0.3\right] e^{i[2\pi (0.1),2\pi (0.4)]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.4,0.5\right] e^{i[2\pi (0.2),2\pi (0.4)]} \\ \,\left[ 0.3,0.5\right] e^{i[2\pi (0.1),2\pi (0.2)]} \end{array} \right) \) | \( \left( \begin{array}{c} \,\left[ 0.4,0.5\right] e^{i[2\pi (0.2),2\pi (0.4)]} \\ \,\left[ 0.3,0.5\right] e^{i[2\pi (0.1),2\pi (0.2)]} \end{array} \right) \) |
Table 4
Complex interval-valued intuitionistic fuzzy decision matrix by \(E_{3}\)
\(\hbox {A}_{{1}}\) | \(\left( \begin{array}{c} \left[ 0.2,0.5\right] e^{i\left[ 2\pi (0.3),2\pi (0.4)\right] } \\ \left[ 0.2,0.3\right] e^{i\left[ 2\pi (0.1),2\pi (0.4)\right] } \end{array} \right) \) | \(\left( \begin{array}{c} \left[ 0.3,0.5\right] e^{i\left[ 2\pi (0.3),2\pi (0.4)\right] } \\ \left[ 0.2,0.3\right] e^{i\left[ 2\pi (0.1),2\pi (0.4)\right] } \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.1,0.3\right] e^{i\left[ 2\pi (0.3),2\pi (0.4)\right] } \\ \left[ 0.3,0.5\right] e^{i\left[ 2\pi (0.1),2\pi (0.2)\right] } \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.4,0.5\right] e^{i\left[ 2\pi (0.2),2\pi (0.4)\right] } \\ \left[ 0.3,0.5\right] e^{i[2\pi (0.1),2\pi (0.2)]} \end{array} \right) \) |
\(\hbox {A}_{{2}}\) | \(\left( \begin{array}{c} \left[ 0.3,0.4\right] e^{i[2\pi (0.2),2\pi (0.5)]} \\ \left[ 0.1,0.4\right] e^{i[2\pi (0.2),2\pi (0.3)]} \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.1,0.4\right] e^{i[2\pi (0.2),2\pi (0.5)]} \\ \left[ 0.2,0.4\right] e^{i[2\pi (0.2),2\pi (0.3)]} \end{array} \right) \) | \(\left( \begin{array}{c} \left[ 0.3,0.4\right] e^{i[2\pi (0.1),2\pi (0.3)]} \\ \left[ 0.1,0.2\right] e^{i[2\pi (0.3),2\pi (0.5)]} \end{array} \right) \) | \(\left( \begin{array}{c} \left[ 0.2,0.4\right] e^{i[2\pi (0.4),2\pi (0.5)]} \\ \left[ 0.1,0.2\right] e^{i[2\pi (0.3),2\pi (0.5)]} \end{array} \right) \) |
\(\hbox {A}_{{3}}\) | \(\left( \begin{array}{c} \left[ 0.2,0.4\right] e^{i[2\pi (0.4),2\pi (0.5)]} \\ \left[ 0.1,0.2\right] e^{i[2\pi (0.3),2\pi (0.5)]} \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.3,0.4\right] e^{i[2\pi (0.4),2\pi (0.5)]} \\ \left[ 0.1,0.2\right] e^{i[2\pi (0.3),2\pi (0.5)]} \end{array} \right) \) | \(\left( \begin{array}{c} \left[ 0.1,0.3\right] e^{i[2\pi (0.3),2\pi \left( 0.4\right) ]} \\ \left[ 0.3,0.4\right] e^{i[2\pi (0.1),2\pi (0.3)]} \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.3,0.4\right] e^{i[2\pi (0.2),2\pi (0.5)]} \\ \left[ 0.1,0.4\right] e^{i[2\pi (0.2),2\pi (0.3)]} \end{array}\right) \) |
\(\hbox {A}_{{4}}\) | \(\left( \begin{array}{c} \left[ 0.4,0.5\right] e^{i[2\pi (0.2),2\pi (0.4)]} \\ \left[ 0.3,0.5\right] e^{i[2\pi (0.1),2\pi (0.2)]} \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.1,0.5\right] e^{i[2\pi (0.2),2\pi (0.4)]} \\ \left[ 0.2,0.5\right] e^{i[2\pi (0.1),2\pi (0.2)]} \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.3,0.4\right] e^{i[2\pi \left( 0.1\right) ,2\pi (0.3)]} \\ \left[ 0.1,0.3\right] e^{i[2\pi \left( 0.3\right) ,2\pi (0.4)]} \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.2,0.5\right] e^{i2\pi [ (0.3),2\pi (0.4)]} \\ \left[ 0.2,0.3\right] e^{i2\pi [ (0.1),2\pi (0.4)]} \end{array}\right) \) |
Table 5
Collective complex interval-valued intuitionistic fuzzy decision matrix
\(\hbox {A}_{{1}}\) | \(\left( \begin{array}{c} \left[ 0.2156,0.4025\right] e^{i\left[ 2\pi (0.2782),2\pi (0.3824)\right] } \\ \left[ 0.1835,0.3656\right] e^{i\left[ 2\pi (0.1188),2\pi (0.2566)\right] }\end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.3526,0.5000\right] e^{i\left[ 2\pi (0.2612),2\pi (0.4000)\right] } \\ \left[ 0.2393,0.3656\right] e^{i\left[ 2\pi (0.1000),2\pi (0.2566)\right] }\end{array} \right) \) | \(\left( \begin{array}{c} \left[ 0.2863,0.4731\right] e^{i\left[ 2\pi (0.2835),2\pi (0.4000)\right] } \\ \left[ 0.2367,0.3608\right] e^{i2\pi (0.1000),2\pi (0.2611)}\end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.2863,0.4731\right] e^{i\left[ 2\pi (0.2835),2\pi (0.4000)\right] } \\ \left[ 0.2367,0.3608\right] e^{i\left[ 2\pi (0.1000),2\pi (0.2611)\right] } \end{array}\right) \) |
\(\hbox {A}_{{2}}\) | \(\left( \begin{array}{c} \left[ 0.2782,0.3824\right] e^{i\left[ 2\pi (0.2156),2\pi (0.4025)\right] } \\ \left[ 0.1188,0.2566\right] e^{i\left[ 2\pi (0.1835),2\pi (0.3656)\right] } \end{array} \right) \) | \(\left( \begin{array}{c} \left[ 0.2357,0.4000\right] e^{i\left[ 2\pi (0.3428),2\pi (0.5000)\right] } \\ \left[ 0.1188,0.2566\right] e^{i\left[ 2\pi (0.2394),2\pi (0.3656)\right] } \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.2835,0.4000\right] e^{i\left[ 2\pi (0.2863),2\pi (0.4731)\right] } \\ \left[ 0.1000,0.2611\right] e^{i\left[ 2\pi (0.2367),2\pi (0.3608)\right] } \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.2835,0.4000\right] e^{i\left[ 2\pi (0.2863),2\pi (0.4731)\right] } \\ \left[ 0.1000,0.2611\right] e^{i\left[ 2\pi (0.2367),2\pi (0.3608)\right] } \end{array} \right) \) |
\(\hbox {A}_{{3}}\) | \(\left( \begin{array}{c} \left[ 0.2156,0.3591\right] e^{i\left[ 2\pi (0.3179),2\pi (0.4192)\right] } \\ \left[ 0.1499,0.2611\right] e^{i\left[ 2\pi (0.1459),2\pi (0.3608)\right] } \end{array} \right) \) | \(\left( \begin{array}{c} \left[ 0.2835,0.4000\right] e^{i\left[ 2\pi (0.3384),2\pi (0.5000)\right] } \\ \left[ 0.1000,0.2611\right] e^{i\left[ 2\pi (0.2367),2\pi (0.3608)\right] } \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.2268,0.3824\right] e^{i\left[ 2\pi (0.3527),2\pi (0.4819)\right] } \\ \left[ 0.1188,0.2566\right] e^{i\left[ 2\pi (0.1835),2\pi (0.3656)\right] } \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.2268,0.3824\right] e^{i\left[ 2\pi (0.3527),2\pi (0.4819)\right] } \\ \left[ 0.1188,0.2566\right] e^{i\left[ 2\pi (0.1835),2\pi (0.3656)\right] } \end{array} \right) \) |
\(\hbox {A}_{{4}}\) | \(\left( \begin{array}{c} \left[ 0.3178,0.4192\right] e^{i\left[ 2\pi (0.2156),2\pi (0.3591)\right] } \\ \left[ 0.1459,0.3608\right] e^{i\left[ 2\pi (0.1499),2\pi (0.2611)\right] } \end{array}\right) \) | \(\left( \begin{array}{c} \left[ 0.1858,0.5000\right] e^{i\left[ 2\pi (0.2638),2\pi (0.4000)\right] } \\ \left[ 0.2156,0.3608\right] e^{i\left[ 2\pi (0.1000),2\pi (0.2611)\right] } \end{array} \right) \) | \(\left( \begin{array}{c} \left[ 0.3527,0.4819\right] e^{i\left[ 2\pi (0.2268),2\pi (0.3824)\right] } \\ \left[ 0.1835,0.3656\right] e^{i\left[ 2\pi (0.1188),2\pi (0.2566)\right] } \end{array} \right) \) | \(\left( \begin{array}{c} \left[ 0.3527,0.4819\right] e^{i\left[ 2\pi (0.2268),2\pi (0.3824)\right] } \\ \left[ 0.1835,0.3656\right] e^{i\left[ 2\pi (0.1188),2\pi (0.2566)\right] } \end{array}\right) \) |
By Using CIIFAAWA operator
Step 1. In this step, we construct the complex interval-valued intuitionistic fuzzy decision matrices and are presented in Tables
2,
3, and
4.
Step 2. Since all the attributes \(A_{i}\) \((i=1,2,3,4)\) are of the benefit type, there is no need for normalization.
Step 3. By utilizing the
CIIFAAWA operator to aggregate all the preference values presented by the experts and the collective complex intuitionistic fuzzy information is presented in Table
5.
Step 4. Utilizing the
CIIFAAWA operator to aggregate all attributes
\(CI_{j}(j=1,2,3,4)\) of the ith line and obtain the overall performance values concerning alternative
\(A_{i}\). Here we take
\(k=3\). Then,
$$\begin{aligned} A_{1}= & {} CIIFAAWA(CIIF_{11},CIIF_{12},CIIF_{13},CIIF_{14}) \\= & {} \left\langle [0.2967,0.4674]e^{i2\pi [0.2768,0.3960]},[0.2211,0.3631]\right. \\ {}{} & {} \left. e^{i2\pi [0.1039,0.2588]}\right\rangle ,\\ A_{2}= & {} CIIFAAWA(CIIF_{21},CIIF_{22},CIIF_{23},CIIF_{24}) \\= & {} \left\langle [0.2718,0.3960]e^{i2\pi [0.2927,0.4674]},[0.1074,0.2211]\right. \\ {}{} & {} \left. e^{i2\pi [0.2588,0.3631]}\right\rangle ,\\ A_{3}= & {} CIIFAAWA(CIIF_{31},CIIF_{32},CIIF_{33},CIIF_{34}) \\= & {} \left\langle [0.2425,0.3821]e^{i2\pi [0.3415,0.4746]}, [0.1191,0.2587]\right. \\ {}{} & {} \left. e^{i2\pi [0.1819,0.3631]}\right\rangle ,\\ A_{4}= & {} CIIFAAWA(CIIF_{41},CIIF_{42},CIIF_{43},CIIF_{44}) \\= & {} \left\langle [0.3199,0.4746]e^{i2\pi [0.2354,0.3821]},[0.1787,0.3631]\right. \\ {}{} & {} \left. e^{i2\pi [0.1191,0.2587]}\right\rangle . \end{aligned}$$
Step 5. Utilizing the score function for ranking the alternatives
\( A_{i}(i=1,2,3,4)\), we have,
$$\begin{aligned} s(A_{1})= & {} 0.1146,s(A_{2})=0.1643,s(A_{3})=0.1458, \\ s(A_{4})= & {} 0.1454 \end{aligned}$$
Step 6. Based on the score values, ranking of the alternatives
\(A_{i}(i=1,2,3,4)\) are
$$\begin{aligned} s(A_{2})>s(A_{3})>s(A_{4})>s(A_{1}) \end{aligned}$$
Therefore, we have
$$\begin{aligned} A_{2}\succ A_{3}\succ A_{4}\succ A_{1}, \end{aligned}$$
where the symbol “
\(\succ \)” means “superior to”. Thus, the best choice is
\(A_{2}\).
By using the CIIFAAWG operator
Here, we apply the CIIFAAWG operator to the above numerical example:
Step \(3'\). Utilizing the CIIFAAWG operator to aggregate all preference value
\(CI_{j}(j=1,2,3,4)\) of the ith line and obtain the overall performance value
\(CI_{i}\) concerning alternative
\(A_{i}\). Here we take
\(k=3\). Then
$$\begin{aligned} A_{1}= & {} CIIFAAWG(CIIF_{11},CIIF_{12},CIIF_{13},CIIF_{14}) \\= & {} \left\langle [0.2749,0.4572]e^{i2\pi [0.2758,0.3954]}, [0.2272,0.3633]\right. \\ {}{} & {} \left. e^{i2\pi [0.1053,0.2589]}\right\rangle , \\ A_{2}= & {} CIIFAAWG(CIIF_{21},CIIF_{22},CIIF_{23},CIIF_{24}) \\= & {} \left\langle [0.2677,0.3954]e^{i2\pi [0.7264,0.5427]},[0.1088,0.2589]\right. \\ {}{} & {} \left. e^{i2\pi [0.2273,0.3633]}\right\rangle ,\\ A_{3}= & {} CIIFAAWG(CIIF_{31},CIIF_{32},CIIF_{33},CIIF_{34}) \\= & {} \left\langle [0.2351,0.3802]e^{i2\pi [0.3396,0.4672]},[0.1246,0.2589]\right. \\ {}{} & {} \left. e^{i2\pi [0.1946,0.3633]}\right\rangle , \\ A_{4}= & {} CIIFAAWG(CIIF_{41},CIIF_{42},CIIF_{43},CIIF_{44}) \\= & {} \left\langle [0.2754,0.4672]e^{i2\pi [0.2317,0.3802]},[0.1867,0.3633]\right. \\ {}{} & {} \left. e^{i2\pi [0.1246,0.2589]}\right\rangle . \end{aligned}$$
Step \(4'\). Utilizing the score function for ranking the alternatives
\( A_{i}(i=1,2,3,4)\), we have,
$$\begin{aligned} s(A_{1})= & {} 0.0995, s(A_{2})=0.1543, s(A_{3})=0.1356, s(A_{4})\\ {}= & {} 0.1144. \end{aligned}$$
Step \(5'\). Based on the score values, ranking of the alternatives
\(A_{i}(i=1,2,3,4)\) are:
$$\begin{aligned} s(A_{2})>s(A_{3})>s(A_{4})>s(A_{1}). \end{aligned}$$
Therefore, we have
$$\begin{aligned} A_{2}\succ A_{3}\succ A_{4}\succ A_{1}, \end{aligned}$$
where the symbol “
\(\succ ''\) means “superior to”. So according to
CIIFAAWG operator
\(A_{2}\) is the best choice.
Discussion about the influence of the parameter k
To reflect the effect of changing, we utilise a separate parameter
k in our suggested technique to rank the alternatives based on the value of the parameter
k. Table
6 displays the ranking results.
Table
6 shows that when the parameter
k is changed, the ranking order derived from the CIIFAAWA operator in each parameter
k remains the same in this example. Also, as the parameter
k for the same alternative goes up, the score function values made by the CIIFAAWA operator go up.
Table 6
Score values and ranking of alternatives for different value of k
\(k=2\) | 0.1143 | 0.1437 | 0.0167 | 0.0959 | \(A_{2}\succ A_{1}\succ A_{4}\succ A_{3}\) |
\(k=3\) | 0.1146 | 0.1643 | 0.1458 | 0.1454 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
\(k=4\) | 0.1172 | 0.1889 | 0.0739 | 0.1057 | \(A_{2}\succ A_{1}\succ A_{4}\succ A_{3}\) |
\(k=5\) | 0.1134 | 0.1872 | 0.0660 | 0.1027 | \(A_{2}\succ A_{1}\succ A_{4}\succ A_{3}\) |
Comparison with different existing methods
This section compares the proposed method with existing methods under the CIIF, CIF, and IIF environments.
Comparison with different CIIF methods
This subsection compares the proposed method to previously developed methods [
52]. The proposed method is compared with the CIIFWG, CIIFWA, CIIFOWA, and CIIFOWG operators [
52]. The score values and ranking of alternatives are presented in Table
7.
Table 7
Comparison of the proposed method with the operators under CIIF environment
CIIFAAWA | 0.1146 | 0.1643 | 0.1458 | 0.1454 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
CIIFAAWG | 0.0995 | 0.1543 | 0.1356 | 0.1144 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
CIIFAAOWA | 0.1143 | 0.1651 | 0.1442 | 0.1391 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
CIIFAAOWG | 0.1292 | 0.1587 | 0.1370 | 0.1143 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| \(-0.0167\) | 0.0203 | 0.0732 | 0.1649 | \(A_{4}\succ A_{3}\succ A_{2}\succ A_{1}\) |
| 0.0225 | \(-0.1750\) | 0.0178 | 0.0953 | \(A_{4}\succ A_{1}\succ A_{3}\succ A_{2}\) |
| 0.1072 | 0.1629 | 0.1379 | 0.1357 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.0831 | 0.1678 | 0.1319 | 0.1261 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
Table
7 shows that by utilizing the proposed CIIFAAWA operator, CIIFAAWG operator, CIIFAAOWG operator, and CIIFAAOWA operator, the same best alternative
\(A_{2}\) is obtained. The ranking result of alternatives by using the CIIFWA operator [
52] and CIIFWG operator [
52] is
\(A_{4}>A_{1}>A_{3}>A_{2}\) and
\(A_{4}>A_{3}>A_{1}>A_{1}\), respectively. The proposed operators are more flexible than existing operators and the existing operators are special cases of the developed operators. In addition, the ranking order of the alternatives by utilizing the CIIFOWA operator [
52] and CIIFOWG operator [
52] are
\(A_{2}>A_{3}>A_{4}>A_{1}\) and
\(A_{2}>A_{3}>A_{4}>A_{1}\) which is the same as obtained by developed operators. However, the proposed operators show more flexibility than the existing operators. But based on changing decision conditions, the above table shows that the best alternatives change with various operators. Hence the present method is more reliable and valid as compared with the existing methods.
Comparison with different CIF methods
To evaluate the performance of the developed method a comparative study was conducted with the methods developed in [
45] and [
51]. The score values and ranking of alternatives are presented in Table
8.
Table 8
Comparison of the proposed method with the operators under CIF environment
| 0.1099 | 0.1173 | 0.1263 | 0.1188 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.0675 | 0.1279 | 0.1218 | 0.1154 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.1343 | 0.1689 | 0.1745 | 0.1390 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.0577 | 0.1009 | 0.1237 | 0.1174 | \(A_{2}\succ A_{4}\succ A_{2}\succ A_{1}\) |
| 0.2308 | 0.2349 | 0.2412 | 0.2476 | \(A_{4}\succ A_{3}\succ A_{2}\succ A_{1}\) |
| 0.2210 | 0.2275 | 0.2493 | 0.2501 | \(A_{4}\succ A_{3}\succ A_{2}\succ A_{1}\) |
| 0.0946 | 0.1379 | 0.1312 | 0.1197 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.2191 | 0.2288 | 0.2464 | 0.2898 | \(A_{4}\succ A_{3}\succ A_{2}\succ A_{1}\) |
CIIFAAWA | 0.1146 | 0.1643 | 0.1458 | 0.1454 | \(A_{2}\succ A_{4}\succ A_{3}\succ A_{1}\) |
CIIFAAWG | 0.0995 | 0.1543 | 0.1356 | 0.1144 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
CIIFAAOWA | 0.1143 | 0.1651 | 0.1442 | 0.1391 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
CIIFAAOWG | 0.1292 | 0.1587 | 0.1370 | 0.1143 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
From Table
8, it is clear that the method developed in this paper gives
\(A_{2}\) as the best alternative. In addition, when comparing the proposed method with other existing methods, it gives the same alternative,
\(A_{2}\), except for the two existing methods, which are the CIFAAWA operator and CIFAAWG operator
\(A_{4}\). The key benefit of developed operators is that they are more professional. It deals with an MCDM problem with both interrelationships among the criteria. The mathematical example validates the application and the effectiveness of the developed method. But based on changing decision conditions, Table
8 shows that the best alternatives change with various operators. Subsequently, our developed techniques are generally more extensive and adaptable than some prevailing techniques to manage the CIIF MCDM problem.
Comparison with different IIF methods
By taking the phase term as zero, the CIIFS reduced to IIFS. This shows that IIFS is a special case of the CIIFS. Therefore, the developed approach is compared with the methods developed under IIFS environments in this subsection. The score values and ranking of alternatives are presented in Table
9.
Table 9
Comparison of the proposed method with the operators under IIF environment
| 0.0578 | 0.1767 | 0.1824 | 0.0923 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.3457 | 0.4980 | 0.5178 | 0.3290 | \(A_{3}\succ A_{2}\succ A_{1}\succ A_{4}\) |
| 0.0470 | 0.1762 | 0.1249 | 0.1159 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.0872 | 0.1583 | 0.1101 | 0.1090 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.0983 | 0.1778 | 0.1209 | 0.1262 | \(A_{2}\succ A_{4}\succ A_{3}\succ A_{1}\) |
| 0.0291 | 0.1648 | 0.1276 | 0.0961 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.3264 | 0.6078 | 0.4991 | 0.3376 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.3309 | 0.4699 | 0.4991 | 0.3456 | \(A_{3}\succ A_{2}\succ A_{4}\succ A_{1}\) |
CIIFAAWA | 0.1146 | 0.1643 | 0.1458 | 0.1454 | \(A_{2}\succ A_{4}\succ A_{3}\succ A_{1}\) |
CIIFAAWG | 0.0995 | 0.1543 | 0.1356 | 0.1144 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
CIIFAAOWA | 0.1143 | 0.1651 | 0.1442 | 0.1391 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
CIIFAAOWG | 0.1292 | 0.1587 | 0.1370 | 0.1143 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
Table 10
Comparison of the proposed method with the operators under IF environment
| 0.0474 | 0.0898 | 0.0607 | 0.0601 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.0388 | 0.0856 | 0.0589 | 0.0542 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
| 0.0248 | 0.0720 | 0.0510 | 0.0550 | \(A_{2}\succ A_{4}\succ A_{3}\succ A_{1}\) |
| 0.0381 | 0.0940 | 0.0572 | 0.0596 | \(A_{2}\succ A_{4}\succ A_{3}\succ A_{1}\) |
| 0.0474 | 0.0920 | 0.0588 | 0.0599 | \(A_{2}\succ A_{4}\succ A_{3}\succ A_{1}\) |
| 0.0406 | 0.0774 | 0.0589 | 0.0659 | \(A_{2}\succ A_{4}\succ A_{3}\succ A_{1}\) |
| 0.0501 | 0.0783 | 0.0588 | 0.0689 | \(A_{2}\succ A_{4}\succ A_{3}\succ A_{1}\) |
| 0.0390 | 0.0589 | 0.0476 | 0.0499 | \(A_{2}\succ A_{4}\succ A_{3}\succ A_{1}\) |
CIIFAAWA | 0.1146 | 0.1643 | 0.1458 | 0.1454 | \(A_{2}\succ A_{4}\succ A_{3}\succ A_{1}\) |
CIIFAAWG | 0.0995 | 0.1543 | 0.1356 | 0.1144 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
CIIFAAOWA | 0.1143 | 0.1651 | 0.1442 | 0.1391 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
CIIFAAOWG | 0.1292 | 0.1587 | 0.1370 | 0.1143 | \(A_{2}\succ A_{3}\succ A_{4}\succ A_{1}\) |
In Table
9, it is clear that our proposed methods that as CIIFAAWA operator, CIIFAAWG operator, CIIFAAOWG operator and CIIFAAOWA operator have the same and best alternatives, which is
\(A_{2}\). In addition, when we compare our suggested method with another existing method, it gives the same alternative,
\(A_{2}\), except for the two existing methods, IIFAAWA and IIFWG operators, which are
\(A_{4}\). According to Table
9, the ranking of alternatives obtained by using the IIFAAWA operator [
50] operator and IIFWG operator is
\(A_{2}>A_{3}>A_{1}>A_{4}\) and
\(A_{3}>A_{2}>A_{1}>A_{4}\), respectively. In addition, IIFAAWA operator [
50] operator is
\(A_{2}>A_{4}>A_{1}>A_{3}\), while the ranking achieved by using the IIFAAWG operator [
50] operator is
\(A_{2}>A_{4}>A_{1}>A_{3}\) which is the same ranking as the CIIFAAWA operator and CIIFAAWG operator. The fundamental reason is that present operators, but not all, examine interrelationships among criteria.
Comparison with different IF methods
In this part, we compare the provided MCDM technique to other existing methods to assess the performance of our suggested model. First, we convert the expert’s decision matrix into IFNs by taking phase terms equal to zero of each CIFN. We now utilize this information by applying the existing aggregation techniques to it. To check the application and effectiveness of the proposed method, we compare our approach to the IFAAWA operator [
49], IFAAWG operator [
49], IFAAWOA operator [
49], IFAAOWG operator [
49], IFWG operator [
10] and IFWA operator [
10]. The ranking results are shown in Table
10.
It is clear that our proposed methods, including CIIFAAWA and CIIFAAWG, have the same ranking as other existing methods, including IFWA [
10], IFWG [
10], IFOWA [
10], and IFOWG [
10]. In addition, the ranking result of IFAAWA [
49], IFAAWG [
49], IFAAOWA [
49] and IFAAOWG [
49] are slightly different from our proposed method. The ranking result of IFAAWA [
49], IFAAWG [
49], IFAAOWA [
49] and IFAAOWG [
49] operator is
\(A_{4}\). But based on changing decision conditions, Table
10 shows that the best alternatives change with various operators.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.