Article

Intuitionistic Hesitant Fuzzy Rough Aggregation Operator-Based

EDAS Method and Its Application to Multi-Criteria
Decision-Making Problems
Muhammad Kamraz Khan 1 , Muhammad Sajjad Ali Khan 2 , Kamran 1, * and Ioan-Lucian Popa 3,4, *

1 Department of Mathematics, Islamia College Peshawar, Peshawar 25120, Khyber Pakhtoonkhwa, Pakistan;
[email protected]
2 Department of Mathematics, Khushal Khan Khattak University, Karak 27200, Khyber Pakhtunkhwa, Pakistan;
[email protected]
3 Department of Computing, Mathematics and Electronics, “1 Decembrie 1918” University of Alba Iulia,
510009 Alba Iulia, Romania
4 Faculty of Mathematics and Computer Science, Transilvania University of Brasov, Iuliu Maniu Street 50,
500091 Brasov, Romania
* Correspondence: [email protected] (K.); [email protected] (I.-L.P.)

Abstract: The fundamental notions of the intuitionistic hesitant fuzzy set (IHFS) and rough
set (RS) are general mathematical tools that may easily manage imprecise and uncertain
information. The EDAS (Evaluation based on Distance from Average Solution) approach
has an important role in decision-making (DM) problems, particularly in multi-attribute
group decision-making (MAGDM) scenarios, where there are many conflicting criteria. This
paper aims to introduce the IHFR-EDAS approach, which utilizes the IHF rough averaging
aggregation operator. The aggregation operator is crucial for aggregating intuitionistic
hesitant fuzzy numbers into a cohesive component. Additionally, we introduce the concepts
of the IHF rough weighted averaging (IHFRWA) operator. For the proposed operator, a new
accuracy function (AF) and score function (SF) are established. Subsequently, the suggested
approach is used to show the IHFR-EDAS model for MAGDM and its stepwise procedure.
Academic Editors: Diego
In conclusion, a numerical example of the constructed model is demonstrated, and a general
García-Zamora and Juan comparison between the investigated models and the current methods demonstrates that
Martínez-Moreno the investigated models are more feasible and efficient than the present methods.
Received: 5 December 2024
Revised: 27 December 2024 Keywords: intuitionistic hesitant fuzzy set; rough set; weighted averaging operator; EDAS
Accepted: 28 December 2024 approach; MAGDM
Published: 30 December 2024

Citation: Khan, M.K.; Khan, M.S.A.; MSC: 94D05; 03B52; 03E72

Kamran; Popa, I.-L. Intuitionistic
Hesitant Fuzzy Rough Aggregation
Operator-Based EDAS Method and Its
Application to Multi-Criteria 1. Introduction
Decision-Making Problems. Axioms
The complexity of the socioeconomic environment increases the complexity of DM
2025, 14, 21. https://doi.org/
10.3390/axioms14010021
challenges. Therefore, under these scenarios, it is difficult for a single decision-maker to
achieve a factual and appropriate result. In the actual world, combining the expertise of
Copyright: © 2024 by the authors.
multiple professional experts is crucial to using DM models to produce more feasible and
Licensee MDPI, Basel, Switzerland.
This article is an open access article
appropriate options. Therefore, in order to generate accurate and fulfilling DM results,
distributed under the terms and MADM offers a highly possible and regulated approach to construct and estimate many
conditions of the Creative Commons competing criteria in every DM domain. A new approach to MAGDM called EDAS uses
Attribution (CC BY) license the alternatives’ deviations from their average scores. The EDAS approach evaluates the
(https://creativecommons.org/
options by utilizing the average solution (AS). Keshavarz Ghorabaee et al. [1] developed it.
licenses/by/4.0/).

Axioms 2025, 14, 21 https://doi.org/10.3390/axioms14010021

Axioms 2025, 14, 21 2 of 24

The approach takes into account two metrics for the appraisal: PDAS (Positive Distance
from Average Solution) and NDAS (Negative Distance from Average Solution). When
our requirements are in conflict, this strategy comes in very helpful. Finding the distance
from the positive ideal solution (PIS) and negative ideal solution (NIS) gives the optimum
alternative in compromise MADM approaches like TOPSIS and VIKOR: the distance
between the PIS and NIS is used to determine which alternative is more desirable. In
these MADM approaches, the optimal option has the highest distance between the PIS
and NIS. The PDAS and NDAS are the two distances from the AS from which the ideal
alternative is chosen in the EDAS technique. According to Keshavarz Ghorabaee et al. [1],
a superior option possesses lower NDAS and greater PDAS values. Zadeh [2] looked at the
widely accepted idea of fuzzy sets (FSs), which effectively handle this kind of imprecise
information, to solve that problem. A membership grade (MG), having a membership
value between 0 and 1, is used to represent fuzzy set information. Since its conception,
this idea has been significantly expanded in a variety of ways, encompassing theoretical
and practical aspects. After that, Atanassov [3] explored the common notion of an IFS,
which is defined by two functions: non-membership grade (non-MG) and MG. The sum
of the MG and non-MG function values for IFSs is constrained to the interval [0, 1]. Since
their establishment, IFSs have drawn a lot of attention from academics who have taken a
diverse approach to studying its hybrid structure. By introducing the notion of an HFS,
Torra [4] developed the idea of an FS. HFS theory allows for a variety of different values to
be possible in situations where it is unclear whether to include an element in a set. Ref. [5]
proposes the development of IHFSs, which are the result of combining IFSs with HFSs. This
idea is useful for dealing with unclear situations, when certain values can represent MGs
and non-MGs for a particular element at the same time. An IHFS is a sophisticated extension
of IFS theory, developed to better represent real-world uncertainty in situations where
decision-makers are uncertain about multiple possibilities for membership. Traditional
IFSs accommodate MG and non-MG, but they assume a single, deterministic value for each.
In contrast, IHFSs account for multiple potential membership values and incorporate them
into the decision framework. IHFSs capture ambiguity more effectively than traditional
fuzzy sets and IFSs by accommodating hesitation in membership definitions, which is
especially useful in complex MADM problems. Pawlak [6] proposed the theory of rough
sets (RSs), in which equivalence relations are essential for determining a subset’s lower (LA)
and upper (UA) approximation. Rough set theory is a granular approach to data analysis,
particularly suited for handling imprecise, inconsistent, or incomplete data. By dividing
data into lower and upper approximations, RSs provide a mechanism to isolate certainties
(lower approximations) and possibilities (upper approximations), leaving ambiguous cases
in the boundary region. It provides a framework for approximating sets when only partial
or imprecise information is available. RSs are particularly effective in situations where the
available data cannot be precisely described due to the absence of complete knowledge.
Further exploration of these theories can be found in Yao [7], Xu and Xia [8], Wei [9],
Qian et al. [10], and Pawlak [11]. Pawlak’s RS has been extended in a number of studies,
including the FRS and RFS proposed by Dubois and Prade [12] and contributions from
other experts in the field [13–17]. Important advantages and insights that are not entirely
achievable in classical settings can be gained via RS analysis in a fuzzy setting. Lower
and upper bounds are used by classical rough sets to approach sets, this helps them to
deal with uncertainty and insufficient knowledge. But they do not clearly distinguish
between MGs and non-MGs. On the other hand, partial membership is supported by
FSs, which more accurately represents many real-world issues which are unclear and
ambiguous. An FS can be estimated in a pristine approximating domain; an FRS expands
on the idea of an RS. This happens in situations when the decision attribute values are
Axioms 2025, 14, 21 3 of 24

fuzzy but the conditional attribute values are exact. Finding a set’s LA and UA is the goal
of FRSs. It is important when the universe of the FS becomes rough, either by converting
the equivalency relation into an appropriate fuzzy connection or by using an equivalency
relation. Chinram et al. [18] introduced the IFRS as a result of the combination of FRSs and
IFSs. Furthermore, with the HFRS, RS methods may now be applied in an HF environment,
with an emphasis on defining the LA and UA in that context. IHFRST, which expands upon
RST, also incorporates the idea of IHFSs. To effectively manage ambiguity, hesitation, and
uncertainty in tasks such as knowledge representation, data processing, and DM, IHFRST
incorporates the ideas of HFRSs, IHFSs, and RSs. In the meantime, because people lack
experience and their thinking is still unclear, professionals may waver between several
assessment values. This is the case with regard to how to create a scientific and effective
choice model to reduce decision chances and clearly comprehend decision outcomes, as
well as how to appropriately and reasonably express the evaluation values of objects after
taking expert understanding and opinions into consideration. By merging the IFRS and
HFRS, we create an IHFRS and go over how it may be used for MADM problems in HF
environments. The IHFRS is an improvement over traditional models that capture the
ambiguity of real-world difficulties by using the LA and UA values with MGs and non-
MGs from the unit interval. The IHFRS expands the information available for problems
involving DM. As far as we are aware, and based on the analysis above, no report of the
application of the EDAS technique with the hybrid studies of IHFSs and RSs employing IHF
averaging in an IHF environment has been made thus far. The IHFR-EDAS model redefines
the capabilities of decision-making frameworks by integrating the nuanced representation
of IHFSs and the granular analysis of RSs into the robust EDAS structure. This innovative
approach delivers precise, adaptable, and transparent decisions, addressing the limitations
of traditional fuzzy- and rough set-based methods. By combining IHFSs and RSs within
the EDAS framework, IHFR-EDAS offers a powerful, nuanced tool for addressing the
multifaceted challenges in modern decision-making scenarios. The IHFR-EDAS model
combines the strengths of IHFSs and RSs to enhance decision-making capabilities. IHFSs
capture nuanced hesitancy and imprecision in stakeholder judgments and RSs refine this
by categorizing uncertain data into precise approximations, reducing noise and ambiguity.
The EDAS model evaluates alternatives based on positive and negative distance from the
average, but integrating IHFSs and RSs allows for a more accurate reflection of uncertainty
and hesitancy in input data. The efficiency of the developed IHF rough EDAS (IHFR-
EDAS) technique has been shown using MAGDM, which is based on IHFR averaging
operators. This drives the current study’s investigation of averaging operators, using the
EDAS method for MAGDM to examine an aggregation operator like IHFRWA.
The remainder of the manuscript is arranged as follows: The basic concepts that are
applied in the subsequent sections are shown in Section 2. In Sections 3 and 4, IFRSs,
HFRSs, and their important properties are defined. We introduce the notion of IHFRSs
with new SFs and AFs for IHFR values (IHFRV) in Section 5. Based on IHFRV, some basic
operations for the suggested concept are shown. Then, the concept of average aggregation
operators, such as IHFRWA, is presented, and Section 6 then goes into detail about their
desirable properties. We introduce the IHFR-EDAS model for MADM based on the notions
addressed in Section 7, and its sequential algorithmic is illustrated, applying the suggested
approach. Section 8 concludes with a numerical illustration using the EDAS approach for
choosing the best small hydro-power plant (SHPP) among Pakistan’s several geographical
areas. Furthermore, a general comparison of the examined models with some current
approaches, comparing to the methods used in Section 9, illustrates that the model under
investigation is more practical and efficient. The paper concludes finally in Section 10.
Axioms 2025, 14, 21 4 of 24

2. Preliminaries
Here, we cover the concepts of FSs, IFSs, HFSs, IHFSs, and RSs, along with their
essential functions, operations, and relationships. These concepts will relate our studies to
the next sections.

Definition 1 ([19]). Let Y be a universal set, then an FS within Y is given as

F = ⟨y, h F (y)|y ∈ Y ⟩, (1)

where h F : Y → [0, 1], and for each y ∈ Y, h F (y) is said to be the MG of y in Y.

Definition 2 ([20]). Let Y be a universal set, then an intuitionistic fuzzy set I within Y is given as

I = { y, h I (y), h′I (y) |y ∈ Y }, (2)

where h I (y) and h′I (y) are functions from Y to [0, 1], ∋ 0 ⪯ h I (y) ⪯ 1, 0 ⪯ h′I (y) ⪯ 1 and
0 ⪯ h I (y) + h′I (y) ⪯ 1, ∀ y ∈ Y.

Definition 3. Let Y be a universal set, then a hesitant fuzzy set H within Y is given as

H = {⟨y, h H (y)⟩|y ∈ Y }, (3)

where h H (y) represents the collection of some values from [0, 1], i.e., the maximum number of the
MG of the element y ∈ Y to H.

Definition 4 ([21]). In the definition of IHFSs, IH in a universal set Y can be defined as follows:

IH = { z, Θ IH (z), Ξ IH (z)|z ∈ Y }, (4)

where Θ IH (z) and Ξ IH (z) are functions from Y to [0, 1], expressing each element z ∈ Y, the
probability of MGs and non-MGs in IH , correspondingly, and for each element z ∈ Y, ∀ u IH (z) ∈
Θ IH (z), ∃v IH (z) ∈ Ξ IH (z) ∋ 0 ≤ u IH (z) + v IH (z) ≤ 1, and ∀ v IH (z) ∈ Ξ IH (z), ∃u IH (z) ∈
Θ IH (z) ∋ 0 ≤ u IH (z) + v IH (z) ≤ 1. IHFN is represented by ŭ = ⟨Θŭ , Ξŭ ⟩ and occurs when Y
has only one element y, Θ IH (z), Ξ IH (z) .

Some important operations of IHFSs are presently articulated as follows:

LetDa = ⟨Φ a , Ψ a ⟩ , a2 = ⟨Φ a2 , Ψ a2 ⟩, and a3E= ⟨Φ a3 , Ψ a3 ⟩ be IHFNs, then
(i) ςa = ∪ha∈Φa {1 − (1 − ha)ξ }, ∪ĥa∈Ψa {ĥa}ξ , ξ ≻ 0;
D E
(ii) aξ = ∪ha∈Φa { ha}ξ , ∪ĥa∈Ψa {1 − (1 − ĥa)ξ } , ξ ≻ 0;
D E
(iii) a2 ∔ a3 = ∪ha2 ∈Φa2 ,ha3 ∈Φa3 {ha2 ∔ ha3 − ha2 .ha3 }, ∪ĥa ∈Ψa ,ĥa ∈Ψa {ĥa2 .ĥa3 } ;
D 2 2 3 3 E
(iv) a2 × a3 = ∪ha2 ∈Φa2 ,ha3 ∈Φa3 {ha2 .ha3 }, ∪ĥa ∈Ψa ,ĥa ∈Ψa {ĥa2 ∔ ĥa3 − ĥa2 .ĥa3 } .
2 2 3 3
Furthermore, suppose that Ẋ ̸= ϕ finite universe. For each N, e M e ∈ I HFS( Ẋ ). The
following is the specification for the N ⊔ M and N ⊓ M of N and M:
e e e e e e

χ, Φ Ne (χ) ⊔ Φ M
e ( χ ), Ψ N
e (χ) ⊓ Ψ M

e⊔M
N e = e (χ) | χ ∈ χ

and
χ, Φ Ne (χ) ⊓ Φ M
e ( χ ), Ψ N
e (χ) ⊔ Ψ M

e⊓M
N e = e (χ) | χ ∈ χ ,
Axioms 2025, 14, 21 5 of 24

n o
− −
where Φ Ne (χ) ⊔ Φ M e (χ) = ν ∈ (Φ N e (χ) ⊔ Φ M e ( χ ))| ν ≥ max( Φ N e ( χ ), Φ Me ( χ )) ,
n o
+ +
Φ Ne (χ) ⊓ Φ M ( χ ) = ν ∈ ( Φ N ( χ ) ∪ Φ M ( χ ))| ν ≤ min ( Φ ( χ ) , Φ ( χ )) and
e e e e e N M
n o
+ +
Ψ Ne (χ) ⊓ Ψ M ( χ ) = ν ∈ ( Ψ N ( χ ) ∪ Ψ M ( χ ))| ν ≤ min ( Ψ ( χ ) , Ψ ( χ )) ,
e
n
e e N
e M
e
o
− −
Ψ Ne (χ) ⊔ Ψ Me (χ) = ν ∈ (Ψ N e (χ) ∪ Ψ M e ( χ ))| ν ≥ max( Ψ e ( χ ), Ψ e ( χ )) . N M

Definition 5 ([20]). Let Y be a universal set and ξ ∈ Y × Y the arbitrary relation on Y, for any
l ∗ ⊑ Y, the LA and UA of l ∗ with respect to the approximation space (AS) (Y, ξ ), defined as

ξ (l ∗ ) = {q∗ ∈ Yξ ∗ (q∗ ) ⊑ l ∗ },

ξ (l ∗ ) = {q∗ ∈ Yξ ∗ (q∗ ) ∩ l ∗ ̸= ϕ}.

Therefore , (ξ (l ∗ ), ξ (l ∗ )) is called a rough set and ξ (l ∗ ), ξ (l ∗ ) : q∗ (Y ) 7→ q∗ (Y ) are the

LA and UA operators.

3. Construction of Intuitionistic Fuzzy Rough Set

This section draws a definition of IFRSs and their important properties.

Definition 6 ([21,22]). Let Y be a universal set and R, ∈ IFS(Y × Y ) be the IF relation. If

σ ⊆ IFS(Y ), the UA and LA of σ with respect to IFAS (Y,R, ) are two IFSs, and represented by
R, (σ ), R, (σ ) and are as follows:
nD E o
R, (σ ) = f , Θ IR(σ) ( f ), Ξ IR(σ) ( f ) / f ∈ Y ,
, ,

nD E o
R, (σ ) = f , Θ IŔ(σ)) ( f ), Ξ IŔ(σ) ( f ) / f ∈ Y ,

where
_h i ^h i
Θ IR(σ) ( f ) = Θ IR, ( f , I ) ∨ Θ Iσ ( I ) ; Ξ IR(σ) (σ ) = Ξ IR, ( f , I ) ∧ Ξ Iσ ( I )
, ,
l ∈U l ∈U

and
^h i _h i
Θ IŔ(v)) ( f ) = Θ IR, ( f , I ) ∧ Θ Iσ ( I ) ; Ξ IŔ(σ) (σ ) = Ξ IR, ( f , I ) ∨ Θ Iσ ( I )
I ∈U I ∈U

such that

0 ≤ (max(Θ IR(σ) ( f ))) + (min(Ξ IR(σ) ( f ))) and 0 ≤ (max(Θ IŔ(σ)) ( f ))) + (min(Ξ IŔ(σ) ( f )))
, ,

Hence, R, (σ ) : R, (σ ) 7→ IFS(Y ) are UA and LA operators, and (R, (σ ), R, (σ )) are IFSs.

The pair
nD E o
R, (σ ) = ( Ŕ(σ ), R, (σ )) = σ, (Θ IR(σ) (σ ), Ξ IR(σ) (σ )), (Θ IŔ(σ)) (ξ ), Ξ IŔ(σ) (σ )) |σ ∈ Y ,
, ,

is called the IFRS. For simplicity,

nD E o
R, (σ ) = σ, (Θ IR(v) (σ ), Ξ IR(σ) (σ )), (Θ IŔσ)) (ξ ), Ξ IŔ(σ) (σ )) |σ ∈ Y ,
, ,

is denoted as R, (σ ) = ((Θ, Ξ), (Ξ, Θ)) and are called IFRS values.
Axioms 2025, 14, 21 6 of 24

Definition 7. The score function for IFRV R, (ξ ) = (R, (ξ ), R, (ξ )) = ((Θ, Ξ), (Ξ, Θ)) is given as

1
2 + Θ + Θ − Ξ − Ξ , Sb(R, (ξ )) ∈ [0, 1].


Sb(R, (ξ )) =
4

Suppose R, (ξ ) = (R, (ξ ), R, (ξ )) = ((Θ, Ξ), (Ξ, Θ)) is an IFRV, then the accuracy function for
a(R, (ξ )) = 14 2 + Θ + Θ + Ξ + Ξ , b


R, (ξ ) is given as follows: b a(R, (ξ )) ∈ [0, 1].

4. Construction of Hesitant Fuzzy Rough Sets

Constructive and deductive approaches to HFRSs were covered by Yang et al. [23].
Even though the axiom technique describes the HFRS in an operator-oriented way, the con-
structive approach uses HF relations to approximate HFSs. We examine their constructive
treatment of tentative HFRSs in the following.

Definition 8. Let Y be a universal set and R b, be an HF relation over Y, ∀ h ∈ HF (Y ); the LA and

b, (h) and bR, (h), respectively, ∀ pe, qe ∈ Y,
UA of h are represented by R

b, / ( pe, qe) ⊻ h(qe)},

b, (h)( pe) = ⊼qe∈Y {R
R

b, (h)( pe) = ⊻qe∈Y {R

R b, ( pe, qe)h(qe)}.

b, , the pair (R
According to the HS relation R b, (h), R
b, (h)) is known as the HFRS of h; both
b, (h) and R
R b, (h) are HFSs. Yang et al. [23] established the subsequent characteristics of
HFR approximations. Regarding a specific universe set Y, R b, is an HF relation over Y , ∀
h ∈ HF (Y ); we have
b, (h)+ ( pe) = min{max{R
(i) R b, + ( pe, qe), (h)+ (qe)}; qe ∈ Y };
(ii) Rb, (h)− ( pe) = min{max{R b, − ( pe, qe), (h)− (qe)}; qe ∈ Y };
b, + ( pe, qe), (h)+ (qe)}; qe ∈ Y };
b, (h)+ ( pe) = max{min{R
(iii) R
(iv) R b, − ( pe, qe), (h)− (qe)}; qe ∈ Y }.
b, (h)− ( pe) = max{min{R

5. Construction of Intuitionistic Hesitant Fuzzy Rough Set

In this section, we draw a definition of IHFRSs and present the counter example for
the IHFRS concept.

Definition 9. Consider Y is the universal set and R, ∈ I HFS(Y × Y ) is the IHF relation. The pair
(Y, R, ) is said to be an IHFAS. If ξ ⊆ I HFS(Y ), the UA and LA of ξ with respect to the IHFAS
(Y, R, ) are two IHFSs, which are represented by R, (ξ ), R, (ξ ) and defined as
nD E o
R, (ξ ) = f , Θ IR(ξ ) ( f ), Ξ IR(ξ ) ( f ) / f ∈ Y ,
, ,
nD E o
R, (ξ ) = f , Θ IŔ(ξ )) ( f ), Ξ IŔ(v) ( f ) / f ∈ Y ,

where
_h i ^h i
Θ IR(ξ ) ( f ) = Θ IR, ( f , I ) ∨ Θ Iξ ( I ) ; Ξ IR(ξ ) (ξ ) = Ξ IR, ( f , I ) ∧ Ξ Iς ( I )
, ,
I ∈U l ∈U

and
^h i _h i
Θ IŔ(ξ )) ( f ) = Θ IR, ( f , I ) ∧ Θ Iξ ( I ) ; Ξ IŔ(ξ ) (ξ ) = Ξ IR, ( f , I ) ∨ Θ Iξ ( I )
I ∈U l ∈U
Axioms 2025, 14, 21 7 of 24

such that

0 ≤ (max(Θ IR(ξ ) ( f ))) + (min(Ξ IR(ξ ) ( f ))) and 0 ≤ (max(Θ IŔ(ξ )) ( f ))) + (min(Ξ IŔ(ξ ) ( f )))
, ,

The pair
nD E o
R, (ξ ) = (R, (ξ ), R, (ξ )) = ξ, (Θ IR(ξ ) (ξ ), Ξ IR(ξ ) (ξ )), (Θ IŔ(ξ )) (ξ ), Ξ IŔ(ξ ) (ξ )) |ξ ∈ Y ,
, ,

is called an IHFRS. For simplicity,

nD E o
R, (ξ ) = ξ, (Θ IR(ξ ) (ξ ), Ξ IR(ξ ) (ξ )), (Θ IŔ(ξ )) (ξ ), Ξ IŔ(v) (ξ )) |ξ ∈ Y ,
, ,

is denoted as R, (ξ ) = ((Θ, Ξ), (Ξ, Θ)) and are called IHFRS values.

Definition 10. The score function for the IHFRV R, (ξ ) = (R, (ξ ), R, (ξ )) = ((Θ, Ξ), (Ξ, Θ)) is
given as
" ! !#
t t
1 1 1
S(R, (ξ )) =
b
4
2+
L
∑ Θj + Θj −
L
∑ Ξj + Ξj , Sb(R, (ξ )) ∈ [0, 1].
j =1 j =1
b b

Suppose R, (ξ ) = (R, (ξ ), R, (ξ )) = ((Θ, Ξ), (Ξ, Θ)) is an IHFRV, then the accuracy function
for R, (ξ ) is as given below:
" #
t t
1 1 1
b(R, (ξ )) =
A
4 b ∑ Θi + Θi + bI ∑ Ξ j + Ξ j , Sb(R, (ξ )) ∈ [0, 1]
I j =1 j =1

For comparing two or more IHFRVs, we apply a score function. Supremacy is indicated
by higher IHFRV scores, and inferiority is shown by lower IHFRV values. If the score
values are equal, we apply the AF.

Proposition 1. Suppose (Y,R, ) is the AS and R, (ξ 1 ) = (R, (ξ 1 ), R, (ξ 1 )) and R, (ξ 2 ) =

(R, (ξ 2 ), R, (ξ 2 )) are two IHFRVs over Y. The following results are satisfied:

(i) ∽ (∽R, (ξ 1 )) = ξ 1 , where ∽R, (ξ 1 ) is the complement of R, (ξ 1 ).

(ii) R, (ξ 1 )⊔R, (ξ 2 ) =R, (ξ 2 )⊔R, (ξ 1 ) and R, (ξ 1 )⊓R, (ξ 2 ) =R, (ξ 2 )⊓R, (ξ 1 ).
(iii) ∽ (R, (ξ 1 )⊓R, (ξ 2 )) = (∽R, (ξ 1 )) ⊔ (∽R, (ξ 2 )).
(iv) ∽ (R, (ξ 1 )⊔R, (ξ 2 )) = (∽R, (ξ 1 )) ⊓ (∽R, (ξ 2 )).
(v) If ξ 1 ⊑ ξ 2 , then R, (ξ 1 ) ⊑R, (ξ 2 ).
(vi) R, (ξ 1 ⊔ ξ 2 ) ⊇R, (ξ 1 )⊔R, (ξ 2 ).
(vii) R, (ξ 1 ⊓ ξ 2 ) ⊆R, (ξ 1 )⊓R, (ξ 2 ).
Now, we present the counterexample for a better explanation of the above concept
of IHFRSs.

Example 1. Let Y = {q1∗ , q2∗ , q3∗ , q4∗ } be an arbitrary set and (Y,R, ) be an IHFAS with R, ∈ Y × Y
the IHF relation presented in Table 1. The best possible normal decision object l ∗ , which is an IHFS,
is now presented by a decision expert and is as described below:

q1∗ , {{0.2, 0.4, 0.6}, {0.1, 0.3, 0.4}} ,

 

 

 ⟨q∗ , {{0.1, 0.4, 0.5}, {0.2, 0.4, 0.5}}⟩,
 

l∗ = 2

 q3∗ , {{0.3, 0.4, 0.5}, {0.2, 0.3, 0.4}} , 

q4∗ , {{0.4, 0.5, 0.6}, {0.1, 0.2, 0.4}}

 

Axioms 2025, 14, 21 8 of 24

Table 1. IHF relation from Y to Y.

ξ q1∗ q2∗ q3∗ q4∗

       
q1∗ {0.2, 0.3, 0.5}, {0.1, 0.3, 0.6}, {0.1, 0.4, 0.5}, {0.1, 0.2, 0.6},
{0.1, 0.3, 0.4} {0.2, 0.3, 0.4} {0.1, 0.3, 0.5} {0.1, 0.3, 0.4}
       
q2∗ {0.1, 0.3, 0.5}, {0.2, 0.3, 0.6}, {0.2, 0.4, 0.5}, {0.1, 0.3, 0.6},
{0.1, 0.2, 0.5} {0.1, 0.2, 0.4} {0.1, 0.4, 0.5} {0.1, 0.2, 0.4}
       
q3∗ {0.2, 0.3, 0.6}, {0.1, 0.2, 0.4}, {0.2, 0.4, 0.5}, {0.1, 0.4, 0.6},
{0.1, 0.3, 0.5} {0.1, 0.2, 0.3} {0.2, 0.3, 0.5} {0.1, 0.3, 0.4}
       
q4∗ {0.1, 0.4, 0.5}, {0.3, 0.4, 0.5}, {0.3, 0.4, 0.5}, {0.2, 0.3, 0.5},
{0.2, 0.3, 0.4} {0.2, 0.4} {0.3, 0.4} {0.1, 0.3, 0.5}

This allows us to use Definition 6 to obtain R, (ξ ) and R, (ξ ).

_h i
Θ IR(ξ ) (q1∗ ) = Θ IR, (q1∗ , l ) ∨ Θ Iξ (l )
,
l ∈U

= (0.2 ∨ 0.2) ∨ (0.4 ∨ 0.3) ∨ (0.6 ∨ 0.5) ∨

(0.1 ∨ 0.1) ∨ (0.4 ∨ 0.3) ∨ (0.5 ∨ 0.6)
∨(0.3 ∨ 0.1) ∨ (0.4 ∨ 0.4) ∨ (0.5 ∨ 0.5) ∨
(0.4 ∨ 0.1) ∨ (0.5 ∨ 0.2) ∨ (0.6 ∨ 0.6)

Θ IR(ξ ) (q1∗ ) = 0.6

,

^h i
Ξ IR(ξ ) (ξ ) = Ξ IR, (q∗ , l ) ∧ Ξ Iς (l )
,
l ∈U

q1∗ , {{0.2, 0.4, 0.6}, {0.1, 0.3, 0.4}} ,

 

 

 ⟨q∗ , {{0.1, 0.4, 0.5}, {0.2, 0.4, 0.5}}⟩,
 

l∗ = 2
∗
 3 , {{0.3, 0.4, 0.5}, {0.2, 0.3, 0.4}} ,
 q 

q4∗ , {{0.4, 0.5, 0.6}, {0.1, 0.2, 0.4}}

 


= (0.1 ∧ 0.1) ∧ (0.3 ∧ 0.3) ∧ (0.4 ∧ 0.4)∧

(0.2 ∧ 0.2) ∧ (0.4 ∧ 0.3) ∧ (0.5 ∧ 0.6)∧
(0.2 ∧ 0.1) ∧ (0.3 ∧ 0.3) ∧ (0.4 ∧ 0.5)∧
(0.1 ∧ 0.1) ∧ (0.3 ∧ 0.2) ∧ (0.4 ∧ 0.4)

Ξ IR(ξ ) (q1∗ ) = 0.1

,

Similarly, in the same manner, we obtained the other values:

Θ IR(ξ ) (q1∗ ) = 0.6, Θ IR(ξ ) (q2∗ ) = 0.6, Θ IR(ξ ) (q3∗ ) = 0.6 , Θ IR(ξ ) (q4∗ ) = 0.6,
, , , ,
Ξ IR(ξ ) (q1∗ ) = 0.1, Ξ IR(ξ ) (q2∗ ) = 0.1, Ξ IR(ξ ) (q3∗ ) = 0.1, Ξ IR(ξ ) (q4∗ ) = 0.1
, , , ,
and Θ IŔ(ξ )) (q1∗ ) = 0.1, Θ IŔ(ξ )) (q2∗ ) = 0.1, Θ IŔ(ξ )) (q3∗ ) = 0.1 , Θ IŔ(ξ )) (q4∗ ) = 0.1
Ξ IŔ(ξ ) (q1∗ ) = 0.5, Ξ IŔ(ξ ) (q2∗ ) = 0.5, Ξ IŔ(ξ ) (q3∗ ) = 0.5, Ξ IŔ(ξ ) (q4∗ ) = 0.5
Thus, the UA and LA IHFAS operators are

R, (ξ ) = {⟨q1∗ , {0.6, 0.1}⟩, ⟨q2∗ , {0.6, 0.1}⟩,

⟨q3∗ , {0.6, 0.1}⟩, ⟨q4∗ , {0.6, 0.1}⟩}

and
Axioms 2025, 14, 21 9 of 24

R, (ξ ) = {⟨q1∗ , {0.1, 0.5}⟩, ⟨q2∗ , {0.1, 0.5}⟩,

⟨q3∗ , {0.1, 0.5}⟩, ⟨q4∗ , {0.1, 0.5}⟩}

Therefore,
R, (ξ ) = (R, (ξ ), R, (ξ ))

= {⟨q1∗ , {0.6, 0.1}, {0.1, 0.5}⟩,

⟨q2∗ , {0.6, 0.1}, {0.1, 0.5}⟩,
⟨q3∗ , {0.6, 0.1}, {0.1, 0.5}⟩,
⟨q4∗ , {0.6, 0.1}, {0.1, 0.5}⟩}.

6. Intuitionistic Hesitant Fuzzy Rough Averaging Aggregation Operator

The notion of IHFRAOs is covered in this portion of the manuscript. They can
be obtained by incorporating the concepts of RSs and IHF averaging operators, and
the section further discusses the fundamental properties of the aggregation concepts of
IHFRWA operators.

Intuitionistic Hesitant Fuzzy Rough Weighted Averaging Operator

A comprehensive evaluation of IHFRWA aggregation operators and their desired
properties is the main objective of this subsection.

Definition 11. Let the collection Ŕ(ξ j ) = {( Ŕ(ξ j ), Ŕ(ξ j )); j = 1, 2, . . . , n)} of IHFRVs with
n
weighted vectors ϖ = (ϖ1 , ϖ2 , . . . , ϖn ) ∋ ∑ ϖ j = 1 and 0 ≤ ϖ j ≤ 1. IHFRWA is defined as
j =1

{⊞nj=1 ϖ j Ŕ(ξ j ), ⊞nj=1 ϖ j Ŕ(ξ j )}

[
IHFRWA {R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ n )} =

Theorem 1. Suppose the collection R, (ξ j ) = {(R, (ξ j ), R, (ξ j )); j = 1, 2, . . . , n)} of IHFRVs with

n
weighted vectors ϖ = (ϖ1 , ϖ2 , . . . , ϖn ) ∋ ∑ ϖ j = 1 and 0 ≤ ϖ j ≤ 1, then the IHFRWA
j =1
operator is determined as

{⊞nj=1 ϖ j Ŕ(ξ j ), ⊞nj=1 ϖ j R, (ξ j )}

[
IHFRWA {R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ n )} =
" ! !#
n n n n
1− (1 − Θ j ) , ϖj
(Ξ j ) ϖj
, 1− (1 − Θ j ) ,
ϖj
(Ξ j ) ϖj
S F F F F
= .
j−=1 j−=1 j−=1 j−=1

Proof. The needed proof is obtained by using mathematical induction.

Then, by the defined operational property’s laws, we obtained

 
R, (ξ 1 ) + R, (ξ 2 ) = ( Ŕ(ξ 1 ) + Ŕ(ξ 2 )), (R, (ξ 1 ) + R, (ξ 2 ))

= ( Θ1 + Θ2 − Θ1 Θ2 , Ξ1 Ξ2 ), Θ1 + Θ2 − Θ1 Θ2 , Ξ1 Ξ2



and
ϱ Ŕ(ξ 1 ) = ( Ŕ(ξ 1 ), R, (ξ 1 ))
Axioms 2025, 14, 21 10 of 24

= (1 − (1 − Θ1 ) ϱ , Ξ1 ) ϱ ), 1 − (1 − Θ1 ) ϱ , Ξ1 ) ϱ



Let n = 2, then

{⊞2i=1 ϖi Ŕ(ξ i ), ⊞2i=1 ϖi R, (ξ i )}

[
I HFRWA(R, (ξ 1 ), R, (ξ 2 )) =
" ! !#
2 2 2 2
(1 − Θ i ) , ( Ξi ) (1 − Θ i ) , ( Ξi )
G G G G
= 1− ϖi ϖi
, 1− ϖi ϖi
.
i −=1 i −=1 i −=1 i −=1

The result is valid for n = 2.

The result is valid when n = l, then

{⊞il=1 ϖi Ŕ(ξ i ), ⊞il=1 ϖi R, (ξ i )}.

[
I HFRWA{R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ l )} =

" ! !#
l l l l
(1 − Θ i ) , ( Ξi ) (1 − Θ i ) , ( Ξi )
G G G G
= 1− ϖi ϖi
, 1− ϖi ϖi
.
i −=1 i −=1 i −=1 i −=1

We then establish that the result is valid for n = l + 1.

+1
{⊞il= l +1
[
I HFRWA{R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ l +1 )} = 1 ϖi Ŕ ( ξ i ), ⊞i =1 ϖi R
, ( ξ i )}.

" ! !#
+1
lG +1
lG +1
lG +1
lG
= 1− (1 − Θ i ) ϖi , ( Ξ i ) ϖi , 1 − (1 − Θ i ) ϖi , ( Ξ i ) ϖi
i −=1 i −=1 i −=1 i −=1

For n = K + 1, the intended outcome is thus true. Therefore, the necessary result is true.
∀n ≥ 1.
According to the study above, the IHFRVs are Ŕ(ξ i ) and R, (ξ i ). Thus, by definition 9,
⊞in=1 ϖi Ŕ(ξ i ) and ⊞in=1 ϖi R, (ξ i ) are also IHFRVs. Therefore, IHFRWA {R, (ξ 1 ),R, (ξ 2 ), . . . ,R, (ξ n )}
is also an IHFRV under the IHF approximation space (Y,R, ).
 
q1∗ , {0.2, 0.4, 0.6}, {0.1, 0.3, 0.4}} ,

 

Example 2. Assume the set ξ ⊑ Y = ⟨q2∗ , {{0.1, 0.4, 0.5}, {0.2, 0.4, 0.5}}⟩, with weighted
 q∗ , {{0.3, 0.4, 0.5}, {0.2, 0.3, 0.4}}
 

3
!t
0.1714327166, 0.2725266523,
vector ϖ =
0.2344326885, 0.3216079427

{⊞4i=1 ϖi Ŕ(ζ i ), ⊞4i=1 ϖi Ŕ(ζ i )}.

[
I HFRWA{ Ŕ(ζ 1 ), Ŕ(ζ 2 ), Ŕ(ζ 3 ), Ŕ(ζ 4 )} =

" ! !#
4 4 4 4
(1 − Θ i ) , ( Ξi ) (1 − Θ i ) , ( Ξi )
[ G G G G
= 1− ϖi ϖi
, 1− ϖi ϖi
.
i −=1 i −=1 i −=1 i −=1
Axioms 2025, 14, 21 11 of 24

 
{0.1397944614, 0.1703255451, 0.2050404220,

 0.2297837569, 0.2571208706, 

 
 0.2882040953, 0.3103589381, 0.3348362148, 
 

 0.3626677080, 0.1811890393, 


 0.2102509145, 0.2432952515, 0.2668478943, 

 
 0.2928694989, 0.3224569451, 
 

 0.3023861628, 0.3271464155, 0.3552996619, 


 0.2361715249, 0.2632819190, 

 
 0.2941073558, 0.3160784579, 0.3403527330, 
 

* 0.3679534064, 0.3492303607, 
+
 0.3723279827, 0.3985907616}, {0.1229462035, 
=
 

 0.1499763195, 0.1801460208, 
 

 0.1890817076, 0.2140735788, 0.2419682903, 


 0.2283895561, 0.2521699900, 

0.2787125541, 0.1599303402, 0.1858206338,
 
 
 

 0.2147181208, 0.2232770021, 


 0.2472150009, 0.2256190428, 0.2609272935, 

0.2837049395, 0.3091282407,
 
 
 

 0.1818396478, 0.2070547136, 0.2351985442, 


 0.2435342069, 0.2668478943, 

0.2928694989, 0.2802025656,
 
 
0.3023861628, 0.3271464155}

Theorem 2 introduces several important characteristics of the IHFRWA operator.

Theorem 2. Let the collection R, (ζ l ) = {(R, (ζ l ), R, (ζ l )); l = 1, 2, . . . , m)} of IHFRVs with

m
weighted vectors ϖ = (ϖ1 , ϖ2 , . . . , ϖn )t ∋ ∑ ϖl = 1 and 0 ≤ ϖl ≤ 1, then some of the
l =1
significant characteristics of the IHFRWA operator appear accordingly.

(i) Idempotency: If R, (ζ l ) = ℵ(ζ ), ∀l = 1, 2, . . . , m, where ℵ(ζ ) = (ℵ(ζ ), ℵ(ζ )) =

((r, t), (r, t). Then, I HFRWA(R, (ζ 1 ),R, (ζ 2 ), . . . ,R, (ζ l )) = ℵ(ζ ).
(ii) Boundedness: Suppose (R, (ζ i ))− = (mini Ŕ(ζ i ), maxi R, (ζ i )) and (R, (ζ i ))+ =
(maxi Ŕ(ζ i ), mini R, (ζ i )). Then, (R, (ζ i ))− ≤ I HFRWA(R, (ζ 1 ),R, (ζ 2 ), . . . ,R, (ζ l )) ≤ (R, (ζ i ))+ .
(iii) Monotonicity: Let ℵ(ιi ) = {(ℵ(ιi ), ℵ(ιi )); i = 1, 2, . . . , n} be another collection of
IHFRVs ∋ ℵ(ιi ) ≤ R, (ξ i ) and ℵ(ιi ) ≤ R, (ζ i ). Then,

I HFRWA(ℵ(ι1 ), ℵ(ι2 ), . . . , ℵ(ι n )) ≤ I HFRWA(R, (ζ 1 ), R, (ζ 2 ), . . . , R, (ζ l ).

(iv) Shift invariance: Consider ℵ(ι)={(ℵ(ι), ℵ(ι)) = ((r, t), (R, , t) are the IHFRVs,
then IHFRWA

(R, (ζ 1 ) + ℵ(ι), R, (ζ 2 ) + ℵ(ι), . . . R(ζ n ) + ℵ(ι)) =

I HFRWA( R(ζ 1 ), R(ζ 2 ), . . . , R(ζ n ) + ℵ(ι).

(v) Homogeneity: For any real number ε ≻ 0, then I HFRWA(ε Ŕ(ζ 1 ), ε Ŕ(ζ 2 ), . . . , ε Ŕ(ζ l )
= εI HFRWA(R, (ζ 1 ),R, (ζ 2 ), . . . , R, (ζ n ).
(vi) Commutativity: If R e, (ξ i ) = {(R
e, (ξ i ), R
e, (ξ i )); i = 1, 2, . . . , n)} is the permutation of
R, (ξ i ) = {( Ŕ(ξ i ), R, (ξ i )), then I HFRWA(R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ l ) = I HFRWA(R e, (ξ 1 ), R
e, (ξ 2 ),
...,R e, (ξ l ).
Axioms 2025, 14, 21 12 of 24

Proof. (i) (Idempotency): As, R, (ξ i ) = ℵ(ξ ), ∀i = 1, 2, . . . , n, where ℵ(ξ ) = (ℵ(ξ ), ℵ(ξ )) =

((r, t), (r, t).

{⊞2i=1 ϖi R, (ξ i ), ⊞2i=1 ϖi R, (ξ i )}
[
I HFRWA(R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ l )) =

" ! !#
2 2 2 2
= 1 − ⨿ (1 − Θ i ) , ⨿ ( Ξ i ) ϖi ϖi
, 1 − ⨿ (1 − Θ i ) , ⨿ ( Ξ i ) ϖi ϖi
,
i =1 i =1 i =1 i =1
∀i, R, (ξ i ) = ℵ(ξ ) = (ℵ(ξ ), ℵ(ξ ) = ((r, t), (r, t).

Therefore,
" ! !#
2 2 2 2
= 1 − ⨿ (1 − r i ) , ⨿ ( t i ) ϖi ϖi
, 1 − ⨿ (1 − r i ) , ⨿ ( t i ) ϖi ϖi
i =1 i =1 i =1 i =1



= (1 − r ), t ), 1 − (1 − r ), t

= (ℵ(ξ ), ℵ(ξ )) = ℵ(ξ )

Hence,

I HFRWA(R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ l )) = ℵ(ξ )

(ii) (Boundedness): As

( Ŕ(ξ i ))− = [(min{Θi }, max{Ξi }), (min{Θi }, max{Ξi })], ( Ŕ(ξ i ))+
i i i i
= [(max{Θi }, min{Ξi }), (max{Θi }, min{Ξi })]
i i i i

and

R, (ςi ) = [(Θi , Ξi ), (Θ, Ξi )]. To show that (R, (ξ i ))− ≤ I HFRWA(R, (ξ 1 ),R, (ξ 2 ), . . . ,R, (ξ l ))
≤ (R, (ξ i ))+ .
As ∀i = 1, 2, . . . n, we obtain

min{Θi } ≤ Θi ≤ max{Θi } ⇔ 1 − max{Θi } ≤ 1 − Θi ≤ 1 − min{Θi }

i i i i

n n n
⇔ ⨿(1 − max{Θi })ϖi ≤ ⨿(1 − Θi )ϖi ≤ ⨿(1 − min{Θi })ϖi
i =1 i i =1 i =1 i

n
⇔ (1 − max{Θi }) ≤ ⨿(1 − Θi )ϖi ≤ (1 − min{Θi })
i i =1 i

n
⇔ 1 − (1 − min{Θi }) ≤ 1 − ⨿(1 − Θi )ϖi ≤ 1 − (1 − max{Θi }).
i i =1 i

Hence,
n
min{Θi } ≤ 1 − ⨿(1 − Θi )ϖi ≤ max{Θi } (5)
i i =1 i
Axioms 2025, 14, 21 13 of 24

Next, ∀i = 1, 2, . . . , n, we obtain

min{Ξi } ≤ Ξi ≤ max{Ξi } ⇔ 1 − max{Ξi } ≤ 1 − Ξi ≤ 1 − min{Ξi }

i i i i

n n n
⇔ ⨿(1 − max{Ξ})ϖi ≤ ⨿(1 − Ξi )ϖi ≤ ⨿(1 − min{Ξi })ϖi
i =1 i i =1 i =1 i

n
⇔ (1 − max{Ξi }) ≤ ⨿(1 − Ξi )ϖi ≤ (1 − min{Ξi })
i i =1 i

n
⇔ 1 − (1 − min{Ξi }) ≤ 1 − ⨿(1 − Ξi )ϖi ≤ 1 − (1 − max{Ξi }).
i i =1 i

Hence,
n
min{Ξi } ≤ 1 − ⨿(1 − Ξi )ϖi ≤ max{Ξi } (6)
i i =1 i

In the same way, we can show that

n
min{Θi } ≤ 1 − ⨿(1 − Θi )ϖi ≤ max{Θi } (7)
i i =1 i

and
n
min{Ξi } ≤ 1 − ⨿(1 − Ξi )ϖi ≤ max{Ξi } (8)
i i =1 i

Therefore, from Equations (5)–(8), we obtain

(R, (ξ i ))− ≤ I HFRWA(R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ l )) ≤ (R, (ξ i ))+ . (9)

(iii) (Monotonicity): As ℵ(ιi ) = {(ℵ(ιi ), ℵ(ιi )) = ((ri , ti ), (ri , ti ) and R, (ξ i ) =

{(R, (ξ i ), R, (ξ i )). To prove that ℵ(ιi ) ≤ Ŕ(ξ i ) and ℵ(ιi ) ≤ R, (ξ i )∀ i = 1, 2, . . . , n. So,

n n
ri ≤ Θi ⇒ 1 − Θi ≤ 1 − r i ⇒ ⨿ (1 − Θ i ) ϖi ≤ ⨿ (1 − r i ) ϖi
i =1 i =1
n n
⇒ 1 − ⨿ (1 − r i ) ϖi ≤ 1 − ⨿ (1 − Θ i ) ϖi
i =1 i =1

Next,
n n
ti ≥ Ξi ⇒ ⨿ t i ϖi ≥ ⨿ Ξ i ϖi (10)
i =1 i =1

We can also demonstrate that

n n
1 − ⨿ (1 − r i ) ϖi ≤ 1 − ⨿ (1 − Θ i ) ϖi (11)
i =1 i =1

Next,
n n
⨿ r ϖi ≥ ⨿ Ξ i ϖi (12)
i =1 i =1

Hence, from Equations (9)–(12), we obtain ℵ(ιi ) ≤ R, (ξ i ) and ℵ(ιi ) ≤ R, (ξ i ).

Therefore, I HFRWA(ℵ(ι1 ), ℵ(ι2 ), . . . , ℵ(ι n )) ≤ I HFRWA(R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ l ).
(iv) Since ℵ(ι) = {(ℵ(ι), ℵ(ι)) = ((r, t), (r, t) is the IHFRV and R, (ξ i ) = {(R, (ξ i ), R, (ξ i ))
= [(Θi , Ξi ), (Θ, Ξi )] is the collection of IHFRVs, so
Axioms 2025, 14, 21 14 of 24

R, (ξ 1 ) + ℵ(ι) = [ Ŕ(ξ 1 ) + ℵ(ι), R, (ξ i ) + ℵ(ι))]

= ((1 − (1 − Θi )(1 − r ), Ξ1 t), (1 − (1 − Θi )(1 − r ), Ξi t)

Therefore,
" #
n
M n
M
(R, (ξ 1 ) + ℵ(ι), R, (ξ 2 ) + ℵ(ι), . . . , R, (ξ n ) + ℵ(ι)) = ϑi (R, (ξ i ) + ℵ(ξ ), ϑi (R, (ξ i ) + ℵ(ξ )
i −1 i −1
  n n
 
1 − ⨿ (1 − Θ i ) ϖi (1 − r i ) ϖi , ⨿ Ξ i ϖi t ϖi ,
 i =1 i =1
 
= n n
 
ϖi ϖi
1 − ⨿ (1 − Θ i ) (1 − r i ) , ⨿ Ξ i t
 ϖ ϖ

i i
i =1 i =1
  n n
 
1 − ( 1 − r i ) ϖi
⨿ ( 1 − Θ i ) ϖi , t
⨿ Ξ i
ϖi ,
i =1 i =1
 
= 
n n
 
ϖ
1 − (1 − r i ) ϖi ⨿ (1 − Θ i ) ϖi , t ⨿ Ξ i i
 
i =1 i =1
  n
 
n

1 − ⨿ (1 − Θ i ⨿ Ξi ) ϖi ,
+ (r, t) ,  ϖi
 i=1 i =1

= n n
  
ϖi
1 − ⨿ (1 − Θ i ) , ⨿ Ξ i
 ϖ

i + (r, t)
i =1 i =1
  n n
 
1 − ⨿ ( 1 − Θ i ) ϖi ,
⨿ Ξ i
ϖi ,
  i =1 i =1
   
= n n
  + (r, t), (r, t)
ϖ
1 − ⨿ (1 − Θ i ) ϖi , ⨿ Ξ i i
 
i =1 i =1

Hence,

(R, (ξ 1 ) + ℵ(ι), R, (ξ 2 ) + ℵ(ι), . . . , R, (ξ n ) + ℵ(ι)) = I HFRWA(R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ n )) + ℵ(ι)

(v) Homogeneity: For any real number ε ≻ 0 and R, , (ξ i ) = {(R, (ξ i ), R, (ξ i )) are

the IHFRVs.
Since
h  i
ε
ε Ŕ(ξ i ) = {(ε Ŕ(ξ i ), εR, (ξ i )) = (1 − (1 − Θi )ε i , Ξi ε i ), 1 − (1 − Θi )ε i , Ξi i

Now,
 
n n
 
1 − ⨿ ( 1 − Θ i ) εϖi , t
⨿ Ξ i
εϖi ,
 i =1 i =1
 
I HFRWA(ε Ŕ(ξ 1 ), ε Ŕ(ξ 2 ), . . . , ε Ŕ(ξ l )) =  n n
 
εϖ
1 − ⨿ (1 − Θi )εϖi , t ⨿ Ξi i
 
i =1 i =1

  ϵ  ϵ 
n n

1 − ⨿ (1 − Θ i ) i , ⨿ Ξ i i
ϖ ϖ ,
 i=1 ϵ i=1
 
= n n
ϵ  
ϖ
1 − ⨿ (1 − Θ i ) ϖi , ⨿ Ξ i i
 
i =1 i =1

= εI HFRWA(R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ n ).

Hence, I HFRWA(ε Ŕ(ξ 1 ), ε Ŕ(ξ 2 ), . . . , ε Ŕ(ξ l ) = εI HFRWA(R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ n ).

Axioms 2025, 14, 21 15 of 24

(vi) Commutativity: Let

h i
I HFRWA(R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ n )) = ⊞in=1 ϖi R, (ξ i ), ⊞in=1 ϖi R, (ξ i )

" ! !#
n n n n
= 1 − ⨿ (1 − Θ i ) ϖi , ⨿ ( Ξ i ) ϖi , 1 − ⨿ (1 − Θ i ) ϖi , ⨿ ( Ξ i ) ϖi .
i =1 i =1 i =1 i =1

Since (R
e, (ξ 1 ), R
e, (ξ 2 ), . . . , R
e, (ξ l )) is any permutation of (R, (ξ 1 ),R, (ξ 2 ), . . . , R, (ξ l )), then

e, (ξ i ), ∀ i = 1, 2, . . . , n.
we obtain R, (ξ i ) = R
" ! !#
n n n ′ n ′
′ ′
= 1 − ⨿ (1 − Θ i ) ϖi , ⨿ ( Ξ i ) ϖi , 1 − ⨿ (1 − Θ i ) ϖi , ⨿ ( Ξ i ) ϖi
i =1 i =1 i =1 i =1
h i
= ⊞in=1 ϖi R
e, (ξ i ), ⊞n ϖi R
i =1
e, (ξ i )

= I HFRWA(R
e, (ξ 1 ), R
e, (ξ 2 ), . . . , R
e, (ξ l )

Hence,

I HFRWA(R, (ξ 1 ), R, (ξ 2 ), . . . , R, (ξ l ) = I HFRWA(R
e, (ξ 1 ), R
e, (ξ 2 ), . . . , R
e, (ξ l )

7. EDAS Approach for MAGDM Utilizing Rough Aggregation Operators

with IHF Information
DM problems become more complex in this competitive environment as the socio-
economic environment becomes more complex. Therefore, it is more difficult for an expert
to make a precise and appropriate decision in this circumstance. In the real world, it is
essential to combine the opinions of a number of competent experts in order to apply
group decision-making models to produce more fulfilling and feasible results. As a result,
MADM has the capacity and discipline to improve and evaluate a variety of competing
criteria in every aspect of decision making to provide more acceptable and workable
decisions. Here, we solve the MADM technique using the EDAS technique. This was
based on the PDAS and NDAS from the AVS. It is thought that the appropriate choice
has a lower NDAS score and a higher PDAS value. To investigate the hybrid structure
of the EDAS technique with IHFRVs, we constructed the IHFR-EDAS method, in which
experts supplied their evaluation results as IHFRVs. The following are the fundamental
stages of construction using the suggested method under IHF rough information. Let the
alternative set m∗ be denoted by q∗ = {q1∗ , q2∗ , . . . , q∗m∗ }, and the decision attribute set n∗
be represented by C ∗ = {c1∗ , c2∗ , . . . , c∗g∗ }. Suppose the professional decision-maker set t∗ is
represented by E∗ = {e1∗ , e2∗ , . . . , et∗∗ }, and they compare their attributes c∗s (s = 1, 2, . . . , g)
to their assessment for each alternative qr∗ (r = 1, 2, . . . , h). Suppose ξ = {ξ 1∗ , ξ 2∗ , . . . , ξ ∗g∗ }
are the weighted vectors for attributes c∗j and ϖ = {ϖ1∗ , ϖ2∗ , . . . , ϖ g∗∗ } are the weighted
n n
vectors for the professional decision-maker set Dl∗ (l = 1, 2, . . . , t) ∋ ∑ χű = 1, ∑ ϑŝ = 1
u =1 v =1
and 0 ≤ χű , ϑŝ ≤ 1. Below is an analysis of the standard algorithm for the EDAS method
with an IHF rough environment.
Combine the expert decision-makers’ evaluations of each alternative qi∗ in relation to
their attribute c∗j , then construct the decision matrix, i.e.,
Axioms 2025, 14, 21 16 of 24

h i
M∗ = R, (ξ ijl )
g∗ × h∗

Where R, (ξ ijl ) indicates the IHFRVs of alternatives qi∗ in relation to their attribute c∗j by
the professional decision-maker set Dl∗ .
Utilizing the proposed approach, by combining the information provided by all
decision-makers against their weight vector, the aggregated decision matrix is produced:

M∗ = R, (ξ ij ) ( g∗ × h∗ )
 

After that, the aggregated matrix is normalized.

R,
h i
R
M∗ = R, (ξ ij ) ( g∗ × h∗ ) to M∗ = R, (ξ ij, )
 
,i.e
( g∗ × h∗ )
   
∗R,
 R, (ξ ij ) = (Θij , Ξij ), (Θij , Ξij ) for benefit 
M =
 R, (ξ n ) = R, (ξ ij )c = (Ξi , Θi ), (Ξi , Θi ) for cost 


ij

Calculate the AVS value for each attribute, applying the recommended method for
every selection.
g∗
" #
1 R
AVS = [AVS](1× h∗ ) = g∗ ∑ R, (ξ ij, )
i =1 (1× h ∗ )

g∗ g∗ g∗ g∗
" ! !#
1 1 1 1
⇒ 1 − ⨿ (1 − Θij ) g∗ , ⨿ (Ξij ) g∗ , 1 − ⨿ (1 − Θij ) , ⨿ (Ξij )
h∗ g∗
i =1 i =1 i =1 i =1 (1× h ∗ )
Step 5: The approach below can be used to calculate the PDAS and NDAS based on
the calculated AVS.
 
R
  max 0, S(R, (ξ ij, )) − S( AVS j )
PDASij = PDASij ( g∗ × h∗ )
= ,
S( AVS j )
 
R
  max 0, S( AVS j ) − S(R, (ξ ij, ))
NDASij = NDASij ( g∗ × h∗ )
=
S( AVS j )
Step 6: The positive weight distance (SPi ) and negative weight distance (SNi ) are
n n
then computed: SPi = ∑ ξ j PDASij and SNi = ∑ ξ j NDASij
j =1 j =1
SPi
Step 7: The (SPi ) and (SNi ) are normalized by applying the formula: ŃSPi = maxi (SPi )
SNi
and ŃSNi = 1 − maxi (SNi )
.
Step 8: Calculate the evaluation score based on ŃSPi and ŃSNi. The appraisal score
( ĀS) value is calculated using the formula that follows:

1
ĀSi = ( ŃSPi + ŃSNi )
2
Step 9: Arrange all of the values in a specific order depending on the ĀSi value. The
greater the ĀSi value, the more beneficial.

8. Example of Using EDAS Method

The efficacy and superiority of the analysed approach is demonstrated by presenting a
real-world MADM example of a small hydro-power plant (SHPP). Consider a construction
company that started a project using the four SHPPs q1∗ , q2∗ , q3∗ , and q4∗ in various loca-
tions throughout Pakistan. To determine which power plant is most suitable for building,
Axioms 2025, 14, 21 17 of 24

these locations are further assessed. These four SHPPs are assessed by the three experts
Ei (i = 1, 2, 3). The experts assess these four SHPPs regarding three criteria: c1∗ = con-
structability, c2∗ = socioeconomic climate, and c3∗ = purchasing and feed-in tariffs, with
ϖ = (0.556040631, 0.1714327168, 0.2725266523)t . In the form of IHFRVs, the competent ex-
perts evaluate each qi∗ ’s assessment report with regard to the relevant criteria. The above
stepwise decision algorithm of the EDAS method is now used along with the generated
IHFRWA operator approach to obtain the most suitable h SHPP
i system.
Step 1: Construct a decision matrix M∗ = R, (ξ ijl ) , which is presented in
m∗ × n∗
Table 2, by collecting the expert decision-makers’ combined evaluation information for
each alternative qi∗ with regard to their criteria ci∗ .

Table 2. IHF rough evaluation information.

ξ c1∗ c2∗ c3∗

     
q1∗ ({0.2, 0.3}, {0.1, 0.4}), ({0.3, 0.4}, {0.1, 0.5}), ({0.1, 0.2}, {0.4, 0.5}),
({0.3, 0.5}, {0.2, 0.3}) ({0.2, 0.5}, {0.1, 0.4}) ({0.2, 0.5}, {0.4, 0.5})
     
q2∗ ({0.3, 0.4}, {0.1, 0.2}), ({0.3, 0.5}, {0.1, 0.2}), ({0.2, 0.5}, {0.3, 0.5}),
({0.1, 0.6}, {0.1, 0.3}) ({0.2, 0.4}, {0.3, 0.5}) ({0.1, 0.4}, {0.3, 0.4})
     
q3∗ ({0.2, 0.4}, {0.3, 0.6}), ({0.3, 0.4}, {0.1, 0.5}), ({0.1, 0.4}, {0.2, 0.5}),
({0.1, 0.5}, {0.1, 0.3}) ({0.3, 0.4}, {0.2, 0.4}) ({0.3, 0.4}, {0.2, 0.3})
     
q4∗ ({0.1, 0.2}, {0.2, 0.4}), ({0.1, 0.2}, {0.3, 0.4}), ({0.2, 0.4}, {0.4, 0.5}),
({0.1, 0.4}, {0.1, 0.2}) ({(0.2, 0.4}, {0.2, 0.5}) ({0.3, 0.4}, {0.4, 0.5})

Step 2: The aggregated decision matrix is M∗ = R, (ξ ij ) ( g∗ × h∗ ) , which is the result of

 

using the IHFRWA operators to aggregate the collective data of decision-makers versus
their weight vector.
Step 3: Although all of the criteria are beneficial, they must be normalized.
Step 4: Applying the suggested methodology, find the value of AVS for each alterna-
tive for every criterion stated in Table 3.

Table 3. The value of the average solution (AVS)of IHFRS.

      

 0.1926000919, 0.1941523420,  
  0.2572475241, 0.2452379293, 

0.2181052989, 0.2332144004, 0.3465445678, 0.2818237031,

     
    


    
 
   

0.2136573611, 0.2713971109, 0.3147464960, 0.2959195844,
   
*

   
 
   
 +
       
0.2959195844, 0.3067148640,   0.3971307547, 0.3300486534,

q1∗
   
 ,   ,  ,



 0.2503771490,  
  0.3568103410,  
  0.3839854317, 
 
 0.2992482541,
0.2740572185, 0.3879878369, 0.4580454741, 0.3332159717,

     
    


    
 
   

0.2740572185, 0.4184634786, 0.4316731951, 0.3463031601,

    
 
   


 
 
 

0.2929900576 0.4466524463 0.5000000001 0.3779899732
   
      

 0.2740572185, 0.1595773808, 
 
 0.1179904378, 0.1950168317, 

0.3613333252, 0.2332144004,  0.2102607332,  0.2281338150, 

     
 

     
   

0.3147464960, 0.1763769039, 0.2102607332, 0.2401363430,

     
   

*
   
+
      
0.3971307547, 0.2485419655,  0.2482643524,  0.2713971109, 
 
q2∗
   
 ,   ,  ,



 0.3336885811,  
  0.2128547996,   0.4381175028,  0.3000000001, 
   
0.4137955371, 0.2818237031,  0.4968981172,  0.3287979790, 

     
 

     
   

0.3710355055, 0.2285893404, 0.4651562602, 0.3392351781,

     
   


 
 
 

0.4466524463 0.2961795986 0.5211082498 0.3664190229
   
Axioms 2025, 14, 21 18 of 24

Table 3. Cont.
      

 0.1926000919, 0.2420975271, 
 
 0.1950168317, 0.1458525123, 

0.2770651944, 0.3332159717, 0.2281338150, 0.1763769039,

     
   


    
   

0.2136573611, 0.3147464960, 0.2160110716, 0.1869555665,

     
   

*
    
   
 +
0.2959195844, 0.3971307547, 0.2482643524, 0.2160110716,
   
q3∗
     
 ,   ,  ,


 0.3119536715,  
  0.4447682480,  
   0.4194429346, 
 0.2572475241,


0.3839327528, 0.5115207068, 0.4433270348, 0.2837910282,

     
   


     
   

0.3298981581, 0.4979901541, 0.4345840639, 0.2929900576,

     
   


 
 
 

0.4000000000 0.5583440359 0.4578452585 0.3182562425
   
      

 0.1284303801, 0.2770651944, 
 
 0.1763769039, 0.2102607332, 

0.1941523420, 0.3121081970,  0.2102607332,  0.2485419655, 

     
 

    
   

0.1458525123, 0.2959195844, 0.2160110716, 0.2713971109,

     
   

*
    
   
+
0.2102607332, 0.3300486534,  0.2482643524,  0.3067148640, 
  
q4∗
   
,   , 
  0.3426228233, , 0.2603251516, 
 


 0.1836823195,  
  0.3839327528,     
0.2452379293, 0.4137955371,  0.3696673005,  0.2961795986, 

     
 

    
   

0.2000000000, 0.4000000000, 0.3742569498, 0.3175858739,

     
   


 
  
  
0.2603251516 0.4290839525 0.4000000000 0.3506647074
  

Step 5: We may determine the score value of the AVS based on the derived AVSi
(i = 1, 2, 3), as shown in Table 3, and then calculate the PDAS and NDAS, as shown in
Tables 4 and 5.

Table 4. The results of PDAS matrix of IHFRS.

c1∗ c2∗ c3∗

q1∗ 0.03828009306 0.03828009306 0.00000000000
q2∗ 0.03836576853 0.000000000000 0.00000000000
q3∗ 0.00000000000 0.02916950736 0.004067812057
q4∗ 0.02591481497 0.00000000000 0.00000000000

Table 5. The results of NDAS matrix of IHFRS.

c1∗ c2∗ c3∗

q1∗ 0.00000000000 0.00000000000 0.09932329278
q2∗ 0.00000000000 0.009375186348 0.08098661866
q3∗ 0.008483035593 0.00000000000 0.00000000000
q4∗ 0.00000000000 0.02603023896 0.02603023896

Step 6: The SPi and SNi are then determined using the criteria weight vector ϖ =
(0.556040631, 0.1714327168, 0.2725266523)t ; it is displayed in Table 6.

Table 6. The results of SPi and SNi of IHFRS.

SP1 0.02784774745 SN1 0.02706824448

SP2 0.02133292614 SN2 0.02367822573
SP3 0.006109195096 SN3 0.004716912464
SP4 0.01440969007 SN4 0.01155636847
Axioms 2025, 14, 21 19 of 24

Step 7: The SPi and SNi should now be normalized as indicated below.

N, SP1 = 1.000000000, N, SP2 = 0.7660557170, N, SP3 = 0.2193784293, N, SP4 = 0.5174454449

and

N, SN1 = 0.000000000, N, SN2 = 0.8747603025, N, SN3 = 0.8257399933, N, SN4 = 0.5730654613

Step 8: The ĀS value is now determined using N, SPi and N, SNi as

ĀS1 = 0.8830278585, ĀS2 = 0.8204080100, ĀS3 = 0.5225592115, ĀS4 = 0.5452554530

Step 9: Table 7 displays the ranking results of the suggested models using the EDAS
technique and are dependent on the calculations described above. Thus, the business
should choose the best SHPP q1∗ . Furthermore, the average of the IHFR data on MGs
and non-MGs is taken. In the form of IFRVs, the combined evaluation of each competent
expert’s qi∗ assessment report with regard to the relevant criteria ci∗ is obtained. The above
stepwise decision algorithm of the EDAS method is now used along with the generated
IFRWA operator approach to obtain the most suitable SHPP system.

Table 7. IHFRS results of ĀSi .

q1∗ q2∗ q3∗ q4∗

0.8830278585 0.8204080100 0.5225592115 0.5452554530

h i
Step 10: Construct a decision matrix M∗ = R, (ξ ijl ) , which is presented in
m∗ × n∗
Table 8, by collecting the combined expert decision-makers’ evaluation information for
each alternative qi∗ with regard to their criteria ci∗ .

Table 8. IF rough evaluation information.

ξ c1∗ c2∗ c3∗

q1∗ ⟨(0.25, 0.25), (0.4, 0.25)⟩ ⟨(0.35, 0.3), (0.35, 0.25)⟩ ⟨(0.15, 0.45), (0.35, 0.45)⟩
q2∗ ⟨(0.35, 0.15), (0.35, 0.2)⟩ ⟨(0.4, 0.15), (0.3, 0.4)⟩ ⟨(0.35, 0.4), (0.25, 0.35)⟩
q3∗ ⟨(0.3, 0.45), (0.3, 0.2)⟩ ⟨(0.35, 0.3), (0.35, 0.3)⟩ ⟨(0.25, 0.35), (0.35, 0.25)⟩
q4∗ ⟨(0.15, 0.3), (0.25, 0.15)⟩ ⟨(0.15, 0.35), (0.3, 0.35)⟩ ⟨(0.3, 0.45), (0.35, 0.45)⟩

The aggregated decision matrix M∗ = R, (ξ ij ) (m∗ × n∗ ) is the result of using the IH-
 

FRWA operators to aggregate the collective data of decision-makers versus their weight
vector. Although all of the criteria are beneficial, they must be normalized.
Applying the suggested methodology, find the value of AVS for each alternative for
each of the criteria listed in Table 9. We may determine the score value of the AVS based
on the derived AVSi (i = 1, 2, 3), as shown in Table 9, and then calculate the PDAS and
NDAS, as shown in Tables 10 and 11.

Table 9. The value of the average solution (AVS) of IFRS.

c1∗ ((0.3380418576, 0.3738952871), (0.4106029333, 0.2583140461))

c2∗ ((0.4004865055, 0.3530916044), (0.4084280166, 0.4106029333))
c3∗ ((0.3380418576, 0.4951145572), (0.4094358073, 0.4716764136))
Axioms 2025, 14, 21 20 of 24

Table 10. The results of PDAS matrix of IFRS.

c1∗ c2∗ c3∗

q1∗ 0.01585899626 0.05123165954 0.00000000000
q2∗ 0.1103575076 0.05123165954 0.03892504292
q3∗ 0.00000000000 0.02678441164 0.1231622086
q4∗ 0.00000000000 0.00000000000 0.00000000000

Table 11. The results of NDAS matrix of IFRS.

c1∗ c2∗ c3∗

q1∗ 0.00000000000 0.00000000000 0.1014702331
q2∗ 0.00000000000 0.00000000000 0.00000000000
q3∗ 0.07863951504 0.00000000000 0.00000000000
q4∗ 0.07863951504 0.1443463237 0.01723306750

The SPi and SNi are then determined using the weight vector of the criteria ϖ =
( 0.556040631, 0.1714327168, 0.2725266523)t , which is displayed in Table 12.

Table 12. The results SPi and SNi of IFRS.

SP1 0.01760102887 SN1 0.02765334293

SP2 0.08075415238 SN2 0.00000000000
SP3 0.03815670886 SN3 0.04372676556
SP4 0.0000000000 SN4 0.07316891818

The SPi and SNi should now be normalized as indicated below.

N, SP1 = 0.2179581898, N, SP2 = 1.000000000, N, SP3 = 0.4725046048, N, SP4 = 0.000000000

and

N, SN1 = 0.000000000, N, SN2 = 0.6220616128, N, SN3 = 0.4023860589, N, SN4 = 1.000000000

The ĀS value is now determined using N, SPi and N, SNi as

ĀS1 = 0.1089790949, ĀS2 = 0.8110308065, ĀS3 = 0.4374453318, ĀS4 = 0.5000000000

Table 13 shows the ranking results of the suggested models using the EDAS approach;
these are dependent on the calculations described above. Thus, the business should choose
the best SHPP q1∗ .

Table 13. The IFRS results of ĀSi .

q1∗ q2∗ q3∗ q4∗

0.1089790949 0.8110308065 0.4374453318 0.5000000000

Table 14 provides the ranking values for the preceding discussion. The advantages of
our proposed approach, as outlined in the above comparative analysis, can be summarized.
The IHFS is well suited for representing uncertain or fuzzy information in MADM problems
Axioms 2025, 14, 21 21 of 24

due to the availability of two sets of MG and Non MG with various possible values, a
feature not achievable by the IFS. The IHFS can also be employed for processing MADM
and comparison methods based on its inherent capabilities. After minor modification of
the IFS, the IHFS uses the general form of the IFS. Because it is capable of supporting
many degrees of MG and Non MG, simultaneously, the IHFS is seen as superior to the
IFS. This enables a fuller and more complex representation of uncertainty and hesitancy.
This results in enhanced DM procedures and more accurate and flexible modeling of
complex problems.

Table 14. Comparison of the IHFRS and IFRS results of ĀSi .

q1∗ q2∗ q3∗ q4∗

0.8830278585 0.8204080100 0.5225592115 0.5452554530
0.1089790949 0.8110308065 0.4374453318 0.5000000000

9. Comparative Analysis
The PDAS and NDAS from the AVS provide the foundation for the EDAS tech-
nique. The suitable option is thought to be the combination of a higher PDAS and
a lower NDAS. Here, a comparison with various current methods in context [18] has
been conducted to establish our investigated IHFR-EDAS approach’s supremacy. Table 7
provides the aggregate results obtained from a comparison analysis of current meth-
ods [18,24–26] with our method, according to Table 2, with criteria weight vector
ϖ = (0.556040631, 0.1714327168, 0.2725266523)t . The developed described example can-
not be solved by applying IHF rough values using the current approaches. Nevertheless,
the methods presented in [27–29] provide approximate information, but they cannot be
used to solve the proposed model. It is obvious that the current methods need certain basic
information and are unable to solve and assess the established instance. Existing meth-
ods like Fuzzy IF-TOPSIS, IF-EDAS, and IF-GRA often focus on single-layer uncertainty
(e.g., membership degrees) while IHFR-EDAS incorporates multi-dimensional hesitancy
(from IHFS) and data-driven approximations (from RS), enabling deeper analysis, and also
IHFR-EDAS provides clearer justifications for rankings by combining subjective hesitancy
evaluations with objective data approximations. Furthermore, existing fuzzy MADM meth-
ods like TOPSIS or AHP often lack nuanced mechanisms to manage both hesitancy and
rough approximations simultaneously and IHFR-EDAS provides dual-layer uncertainty
handling. Traditional fuzzy MADM models (e.g., Fuzzy TOPSIS, Fuzzy AHP) focus on
membership grades, they do not adequately capture multi-valued hesitancy or distinguish
between certain and uncertain classifications while IHFR-EDAS bridges this gap by incor-
porating both hesitant fuzzy parameters and RS-based approximations, offering a richer
representation of uncertainty. Therefore, the suggested approach is more capable and
efficient than the current approaches.

An Analysis Comparing the Current IHFR-EDAS Approach with IFR-EDAS

If the MG and non-MG degrees are both occupied by a single element, IFRSs can be
regarded as a specific example of IHFRSs. Transferring IHFRSs to IFRSs involves finding
out the average values of the MG and non-MG degrees in order to make comparison
easier. Following conversion, Table 8 can display the intuitionistic information. IFR-EDAS
can then be used to calculate various assessment values. The final alternative ranking
is q1∗ ≥ q2∗ ≥ q4∗ ≥ q3∗ and the preferred alternative is q1∗ . It is evident that the ranking
obtained from the method suggested by the suggested technique differs from IFRSs [18].
The main goal is to average the MG and non-MG degrees of the IFRSs, which could lead
Axioms 2025, 14, 21 22 of 24

to information loss and deception. To facilitate a comparison, transforming a IHFRS to a

IFRS involves calculating the average values of MG and Non MG degrees. Once converted,
the intuitionistic information is presented in Table 8. Subsequently, the broad assessment
values can be computed using EDAS within an IF environment.

10. Conclusions
The factual information on specific facts usually remains unknown in DM problems,
and this lack of clarity contributes complexity and difficulty to the DM process. RSs
and IHFS are general mathematical instruments that can easily deal with ambiguous and
imprecise information. The EDAS technique is crucial to the DM process when there
are more conflicting criteria in MADM scenarios. We develop the IHFR-EDAS technique
to investigate a hybrid structure of the EDAS technique using IHFRVs. Introducing the
IHFR-EDAS approach based on the IHF rough averaging operator is the objective of this
work. Furthermore, we propose the idea of IHFRWA operators. A detailed description of
the developed operator’s basic desirable features is provided. For the proposed operators,
new accuracy and scoring functions are constructed. The suggested method is then used to
illustrate the IHFR-EDAS model for MADM and its iterative algorithm. The built model is
finally shown numerically, and a general comparison of the models under study with a few
contemporary methods shows that the models under study are more effective and useful
than the methods now in use. Using intuitionistic, hesitant, and Pythagorean fuzzy data,
we will expand on the suggested approach in subsequent research to incorporate a variety
of aggregation operators, including Dombi operations, Einstein operations, etc. We will
also concentrate on how the suggested approach may be used in various real-world issues
utilizing Pythagorean, hesitant, and intuitionistic fuzzy information. Furthermore, we will
apply the developed method to other domains, including medical diagnostics, and expand
it to other generalizations of FSs. Many real-world problems involve conflicting opinions,
incomplete data, and hesitation. IHFR-EDAS effectively models such scenarios. The EDAS
model helps decision-makers evaluate complex scenarios more reliably, reducing bias
and error, and its application is useful in many areas such as supply chain management,
healthcare, environmental management, multi-objective engineering design, social sciences,
etc. The IHFR-EDAS approach is typically static and may not work well in situations when
preferences, criteria, or alternatives vary over time. The IHFR-EDAS approach might be
overly complicated for small-scale decision-making situations when compared to more
straightforward MADM methods like AHP and TOPSIS. Particularly, when working with
complicated FSs, with a large number of alternatives, or criteria, the IHFR-EDAS approach
can require a substantial amount of processing power. The IHFR-EDAS approach may
not be suitable for simple problems or situations where clear data are adequate, even
when it works effectively for complicated, unclear DM scenarios. The approach works
independently and could be difficult to combine with other frameworks like machine
learning, game theory, or simulation-based techniques. Future research will expand the
suggested approach to include various aggregation operators, such as the Bonferroni mean,
Einstein operations, Maclaurin symmetric mean operators, Hamacher operations, Dombi
operations, the Choquet integral, and interaction aggregation operators.

Author Contributions: Conceptualization, M.K.K. and M.S.A.K.; data curation, M.K.K., M.S.A.K.
and K.; formal analysis, K. and I.-L.P.; funding acquisition, I.-L.P.; investigation, M.K.K., M.S.A.K.
and I.-L.P.; methodology, M.K.K. and M.S.A.K.; project administration, I.-L.P.; software, M.S.A.K.;
supervision, K.; validation, M.S.A.K. and K.; visualization, M.K.K. and K.; writing–original draft,
M.K.K. and M.S.A.K.; writing—review and editing, M.K.K., M.S.A.K., K. and I.-L.P. All authors have
read and agreed to the published version of the manuscript.
Axioms 2025, 14, 21 23 of 24

Funding: This research received no external funding.

Data Availability Statement: Data is contained within the article.

Conflicts of Interest: The authors declare no conflicts of interest.

List of Symbols

FS Fuzzy set
IFS Intuitionistic fuzzy set
HF Hesitant fuzzy set
IHFS Intuitionistic hesitant fuzzy set
RS Rough set
Non-MG Grade of non-membership
⊞ Summation
⨿ Product
W
Maximum
V
Minimum
Ŕ(ξ ) Collection of IHFRVs

References
1. Keshavarz Ghorabaee, M.; Zavadskas, E.K.; Olfat, L.; Turskis, Z. Multi-criteria inventory classification using a new method of
evaluation based on distance from average solution (EDAS). Informatica 2015, 26, 435–451. [CrossRef]
2. Zadeh, L.A. Fuzzy sets. Inf. Control 1965, 8, 338–353. [CrossRef]
3. Atanassov, K.T. Intuitionistic fuzzy sets. Fuzzy Sets Syst. 1986, 20, 87–96. [CrossRef]
4. Torra, V. Hesitant fuzzy sets. Int. J. Intell. Syst. 2010, 25, 529–539. [CrossRef]
5. Peng, J.J.; Wang, J.Q.; Wu, X.H.; Zhang, H.Y.; Chen, X.H. The fuzzy cross-entropy for intuitionistic hesitant fuzzy sets and their
application in multi-criteria decision-making. Int. Syst. Sci. 2015, 46, 2335–2350. [CrossRef]
6. Pawlak, Z. Rough sets. Int. J. Comput. Inf. Sci. 1982, 11, 341–356. [CrossRef]
7. Yao, Y.Y. Constructive and algebraic methods of the theory of rough sets. Inf. Sci. 1998, 109, 21–47. [CrossRef]
8. Xu, Z.; Xia, M. Distance and similarity measures for hesitant fuzzy sets. Inf. Sci. 2011, 181, 2128–2138. [CrossRef]
9. Wei, G. Hesitant fuzzy prioritized operators and their application to multiple attribute decision making. Knowl.-Based Syst. 2012,
31, 176–182. [CrossRef]
10. Qian, G.; Wang, H.; Feng, X. Generalized hesitant fuzzy sets and their application in decision support system. Knowl.-Based Syst.
2013, 37, 357–365. [CrossRef]
11. Pawlak, Z. Rough Sets: Theoretical Aspects of Reasoning About Data; Kluwer Academic: Norwell, MA, USA, 1991.
12. Dubois, D.; Prade, H. Rough fuzzy sets and fuzzy rough sets. Int. J. Gen. Syst. 1990, 17, 191–209. [CrossRef]
13. Som, T.; Castillo, O.; Tiwari, A.K.; Shreevastava, S. Fuzzy, Rough and Intuitionistic Fuzzy Set Approaches for Data Handling; Springer:
Singapore, 2023.
14. Li, T.J.; Zhang, W.X. Rough fuzzy approximation on two 967 universes of discourse. Inf. Sci. 2008, 178, 892–906. [CrossRef]
15. Lin, G.; Qian, Y.; Li, J. NMGRS: Neighborhood-based multigranulation rough sets. Int. J. Approx. Reason. 2012, 53, 1080–1093.
[CrossRef]
16. Zhan, J.; Xu, W. Two types of coverings based multigranulation rough fuzzy sets and applications to decision making. Artif. Intell.
Rev. 2020, 53, 167–198. [CrossRef]
17. Liu, C.; Miao, D.; Qian, J. On multi-granulation covering rough sets. Int. J. Approx. Reason. 2014, 55, 1404–1418. [CrossRef]
18. Chinram, R.; Hussain, A.; Mahmood, T.; Ali, M.I. EDAS method for multi-criteria group decision making based on intuitionistic
fuzzy rough aggregation operators. IEEE Access 2021, 9, 10199–10216. [CrossRef]
19. Xue, Z.; Sun, B.; Hou, H.; Pang, W.; Zhang, Y. Three-way decision models based on multi-granulation rough intuitionistic hesitant
fuzzy sets. Cogn. Comput. 2022, 14, 1859–1880. [CrossRef]
20. Zhou, L.; Wu, W.Z. On generalized intuitionistic fuzzy rough approximation operators. Inf. Sci. 2008, 178, 2448–2465. [CrossRef]
21. Xu, Z.S. A deviation-based approach to IF multiple attribute group decision-making. Group Decis. Negot. 2010, 19, 57–76.
[CrossRef]
22. Xia, M.; Xu, Z. Hesitant fuzzy information aggregation in decision making. Int. J. Approx. 2011, 52, 395–407. [CrossRef]
23. Yang, X.; Song, X.; Qi, Y.; Yang, J. Constructive and axiomatic approaches to hesitant fuzzy rough set. Soft Comput. 2014, 18,
1067–1077. [CrossRef]
Axioms 2025, 14, 21 24 of 24

24. Zhang, Z. Generalized intuitionistic fuzzy rough sets based on intuitionistic fuzzy coverings. Inf. Sci. 2012, 198, 186–206.
[CrossRef]
25. Dong, J.Y.; Wan, S.P. Interval-valued trapezoidal intuitionistic fuzzy generalized aggregation operators and application to
multi-attribute group decision making. Sci. Iran. E 2015, 22, 2702–2715.
26. Wan, S.P.; Wang, Q.Y.; Dong, J.Y. The extended VIKOR method for multi-attribute group decision making with triangular
intuitionistic fuzzy numbers. Knowl.-Based Syst. 2013, 52, 65–77. [CrossRef]
27. Zhang, X.; Zhou, B.; Li, P. A general frame for intuitionistic fuzzy rough sets. Inf. Sci. 2012, 216, 34–49. [CrossRef]
28. Yun, S.M.; Lee, S.J. Intuitionistic fuzzy rough approximation operators. Int. J. Fuzzy Log. Intell. Syst. 2015, 15, 208–215. [CrossRef]
29. Zhang, H.; Shu, L.; Liao, S. Intuitionistic fuzzy soft rough set and its application in decision making. Abstr. Appl. Anal. 2014, 2014,
287314. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.