1. Introduction
Inequality is the comparison between two quantities while in mathematics it signifies inconsistency between the quantities. In many branches of mathematics, including calculus, numerical analysis, functional analysis, and other engineering disciplines, the notion of inequality is widely used. Since mathematical inequalities have a greater influence on error estimates, during the past few decades, a number of authors have studied the error analysis of different Newton–Cotes type quadrature procedures. In the modern sense, inequality is essentially a necessary component of all disciplines of applied mathematics. It is discovered that inequalities have a broad spectrum of applications in several scientific fields, such as the physical and engineering sciences. One of its area of applications is the notion of convexity; therefore there exists a nice link between mathematical inequalities and the theory of convex functions.
Definition 1 ([1]). A function is convex, ifholds ∀
and . The most intriguing aspect of the theory of inequalities is to get the modified or refined inequalities for convex function and these modified inequalities can be determined while working on the refined version of the convex type functions. Several types of the generalization of the convex functions have been obtained and one of the generalization is the superquadratic function.
Abramovich et al. were the first to propose the concept of superquadraticity [
2]. The theory of integral inequalities is significantly improved by superquadraticity, which offers tighter and better limits than general convexity. In applied mathematics applications, where improved approximation results in higher modeling accuracy, and in optimization problems, where the best solution depends on the quality of the boundary estimate, such an enhancement is greatly sought.
If the function
is superquadratic on
, then
holds
, where
. Moreover, the function
is subquadratic, if the inequality (
2) is reversed.
Let us consider a function
defined on
. This function is superquadratic for
, and subquadratic for
. In the case
is set equal to
, equality holds in (
2) at
.
It is to be noted that every superquadratic function fulfills the three extra conditions listed by the following lemma:
Lemma 1. If the function ψ is superquadratic, then
.
for ψ being a differentiable function at and .
ψ is convex, in the case ψ is positive and .
The following lemma explains the relationship between convexity and superquadraticity.
Lemma 2. The function ψ is superquadratic, if is convex and . Its converse is false.
The definition of superquadraticity based on line of support is given in (
2), while an alternative definition was also introduced by Abramovich et al. [
2], stated as follows.
Definition 2. If the function ψ is superquadratic, thenholds for every and . Remark 1. The function ψ is subquadratic, if the inequality sign in (3) is reversed. Two fundamental inequalities that have greatly contributed to the advances in the theory of superquadraticity are the Jensen and -type inequalities. These results are among the most significant and widely applied in the study of superquadratic functions.
Theorem 1. If the function ψ is superquadratic, thenholds , and , where = , and . In the context of superquadraticity, Banić et al. [
3] derived the
-type inequalities presented below.
Theorem 2. Let the function ψ is superquadratic on , then Building on this basis, the authors of [
4] made a substantial expansion of superquadraticity into the field of fractional calculus by investigating the fractional viewpoint of the inequalities of
-type using
fractional integrals.
Theorem 3. Let be a superquadratic function on thenwhere and are given byand The notations and are the left and right sided operators. Where Γ is stated as a gamma function and given by .
Superquadratic functions have been recently explored in depth, particularly regarding their applications to information theory [
5]. Further developments were made by Alomari et al. [
6]. They initiated the notion of h-superquadratic function, investigated its properties and inequalities. In this direction, Mario Krnić and collaborators further contributed to the field by introducing the notion of a multiplicatively superquadratic function in [
7]. Khan et al. [
8] further this line of inquiry by proposing the (P, m)-superquadratic function, a more generic form that enriches the functional landscape of superquadraticity by incorporating instances, important features, integral inequalities, and real-world applications.
Definition 3. If ψ is (P, m)-superquadratic function on for and some fixed , thenholds and . The function ψ is (P, m)-subquadratic, provided is (P, m)-superquadratic. Butt and Khan [
9] proposed a new form of superquadraticity, superquadratic
, and its inequalities and applications in information theory within the framework of fractional interval calculus. For readers interested in convexity and superquadraticity, we refer to [
10,
11,
12,
13,
14], and the references cited in them for a complete understanding of the applications and uses in the framework of inequality theory.
Uncertainty is a major consideration in many practical problems. When these problems are modeled with the use of exact numerical values, there may be significant errors, resulting in the generation of unreliable results. Thus, the challenge of finding effective ways to reduce these errors has remained an important consideration. The idea of interval analysis was first proposed by Moore [
15] in 1969 as a means of automated error evaluation. The method addressed the problem by improving the accuracy of computations and attracting much research interest. In interval analysis, interval numbers are considered as variables, and arithmetic calculations are performed with the use of intervals instead of exact numbers, since intervals contain uncertainty.
Several researchers, including Chalco-Cano et al. [
16], Flores-Franulič et al. [
17], and Costa et al. [
18], have expanded classical inequalities to interval and fuzzy calculi. In particular, Zhao et al. [
19] formulated an
h-convex
and established corresponding inequalities through the interval inclusion relation. Later, in 2021, Khan et al. [
20] defined
h-convex
, by employing the Kulisch-Miranker order, deriving various inequalities for these classes of convex functions.
However, inequalities based on partial order relations, such as fuzzy, inclusion, or pseudo-order relations, tend to be less accurate because two intervals under such relations may not always be comparable. Consequently, identifying an appropriate ordering framework for studying inequalities involving interval-valued functions remains a challenging issue. To address this, Bhunia et al. [
21] introduced the
-order in 2014, establishing a complete ranking relation that enables the comparison of any two intervals. Building upon this idea, Khan and Butt [
22] made a significant contribution by developing
-superquadratic
s and their integer and fractional version inequalities, demonstrating their effectiveness through multiple applications. For additional studies on interval based inequalities, readers may refer to [
23,
24,
25,
26].
We begin by presenting the essential mathematical preliminaries associated with the order relation to formulate the idea of -superquadratic functions.
As stated by Moore [
15], interval numbers generalize the concept of real numbers, while interval arithmetic extends the operations of real arithmetic. An interval number
is defined as follows:
Each real number
can be represented as an interval number
having zero width. Interval numbers can alternatively be represented in the
form. In this representation, an interval
is expressed as follows:
where
is the formula for computing the center and
is the formula for computing the radius of
.
Definition 4 ([24]). The addition, subtraction and multiplication of -intervals are defined as follows: and are defined as follows:
Definition 5 ([21]). Let , then the order relation is defined as follows: For any , either or .
The relation satisfies properties of total order relation for every and :
Theorem 4 ([23]). Let the function be an interval valued such that , then ψ is interval Riemann integrable on if and only if and are Riemann integrable on such that In the framework of the -order relation, the integral maintains the order of intervals.
Theorem 5. Let the functions , be interval valued such that and . If , ∀
, , then This study is distinctive in employing the concepts of (P, m)-superquadratic functions and its fractional counterpart to establish inequalities within the -ordered relation, a direction not previously explored in the literature. It further integrates classical integral inequalities, including Jensen’s, -type, and fractional -type inequalities, into the framework of -interval valued function, thereby providing novel insights and extensions in the field of inequalities. It is noteworthy that the -ordered interval valued analysis fundamentally differs from the conventional interval valued analysis, as intervals here are characterized through their centers and radii.
Motivation Behind Introducing the Concept of
-(P, m)-Superquadratic
In many practical situations involving uncertainty, quantities cannot be described by precise real numbers and are instead represented by intervals. Classical approaches for comparing such intervals often rely on partial order relations, such as inclusion or pseudo-order relations. However, a major limitation of these relations is that two intervals are not always comparable, which makes it difficult to establish meaningful inequality relations. To illustrate this limitation informally, consider the task of finding the home of Mr. A in a street containing 200 houses. If no additional information is available, searching for the exact house would be extremely difficult. If we instead know that Mr. A’s home lies somewhere between house number 10 and house number 120, the problem becomes easier because the search space is reduced to an interval; nevertheless, locating the house within such a wide range may still be challenging. A more informative description would be to specify the location using a center and radius type interval; for example, stating that the house is around the 50th house with a deviation of about 6 houses. This description immediately indicates that the house lies between the 44th and 56th houses, which significantly reduces the search region and makes the task much easier. In a similar way, the center-radius order provides a more informative and effective way to compare intervals than classical partial orders, since it incorporates both the midpoint and the spread of the interval and therefore allows a more structured comparison of uncertain quantities. Motivated by this advantage, the present work integrates the center-radius order framework with the concept of (P, m)-superquadratic functions to obtain refined analytical results. In particular, the definition of (P, m)-superquadratic functions involves the subtraction of two additional terms on the left-hand side, which leads to a more refined approximation compared with classical convexity. Moreover, the parameter m provides additional flexibility in describing the behavior of the function; for instance, by taking m = 1 we recover the results corresponding to P-superquadratic functions. This parameterized structure allows us to capture richer information about the function’s behavior, whereas the classical superquadratic function typically provides information only for a single fixed form. Another advantage is that many results derived under the (P, m)-superquadratic framework remain comparatively simpler and more flexible than those obtained directly from the standard superquadraticity framework. Therefore, the combination of the center-radius order with the (P, m)-superquadratic structure enables the development of sharper and more informative inequalities for interval-valued functions. These refined inequalities further lead to improved divergence measures in information theory, particularly in situations where uncertainty or imprecise data must be modeled through interval-valued quantities.
The structure of the paper is as follows:
The background information and prerequisites for interval valued analysis and associated inequalities are given in
Section 1. With the help of numerical estimations and graphical representations, we present the idea of
-ordered interval valued (P, m)-superquadratic functions in
Section 2. We also establish new interval versions of Jensen and
-type inequalities related to
-superquadratic function. By creating fractional
-type inequalities for
-interval valued superquadratic function,
Section 3 expands these findings to the fractional domain. Numerous examples, graphical representations, and numerical analyses are provided to support the theoretical findings. The information theory applications of the obtained results are highlighted in
Section 4. Lastly,
Section 5 provides closing thoughts and suggests possible paths of inquiry for further study.
2. -(P, m)-Superquadratic and Its Integer Order Inequalities
In this section, we first present the definitions of --superquadratic and -P-superquadratic functions. Building upon these definitions, we introduce the concept of the --superquadratic , discuss its properties in detail, and derive the associated inequalities.
Definition 6. An interval valued function , such that is said to be -m-superquadratic function on = , where and , if the following inequalityholds for any and . We say that ψ is a --subquadratic function if is --superquadratic function. Note that in this scenario is m-superquadratic and is -subquadratic function on . The idea of conventional -superquadratic functions on for is obviously recaptured in this formulation.
Definition 7. An interval valued function , such that , is called -P-superquadratic function on , ifholds ∀
and . Here, is P-subquadratic and is P-superquadratic function on . Remark 2. When the inequality in (11) holds in the opposite direction, the function ψ is called -P-subquadratic. Definition 8. An interval valued function , such that , is called --superquadratic function on , where , ifholds ∀
, and . Remark 3. When the inequality in (12) holds in the opposite direction, the function ψ is called --subquadratic. Remark 4. If is substituted into (12), the definition reduces to that of the -P-superquadratic function. We now present a simple example of
-
-superquadratic and
-
-subquadratic
s to illustrate the underlying concept of
-
-superquadraticity. It should be observed that if
is
-
-subquadratic
, then
becomes
-
-superquadratic. Furthermore, for a
-
-subquadratic
, the direction of the order symbol
in (
12) is reversed.
Example 1. The below mentioned function is a --superquadratic on if and only if , for every and . The center of the interval is , while the radius of the interval is The Graphs a and b given by Figure 1 validate the above mentioned function for by employing Definition 8 for various values of , and ρ. In the subsequent section, we investigate the distinctive characteristics of --superquadratic .
Proposition 1. Let ψ and are --superquadratic s such that and , respectively, then and , where are --superquadratic s.
Proof. Suppose that
and
are considered to be
-
-superquadratic
s, i.e.,
,
and
, we have
or
and
or
Next adding (
14) and (
15), and using the properties of center and radius values, we achieve that
□
Thus, by (
16) and (
17), we infer that
and
are
-
-superquadratic.
Proposition 2. Let ψ is a --superquadratic such that and is a --convex function such that on , then is a --superquadratic .
Proof. Since
and
then we have
By Definition 5, the results (
19) and (
20) can be combined and written as follows:
This demonstrates that is a --superquadratic . □
Proposition 3. If ψ is a non-negative --superquadratic function such that , then ψ is a --superquadratic.
Proof. Let
is a
-
-superquadratic
function, so we have
for any
and
.
By Definition 5, the results (
21) can be written as follows:
and
First consider (
22) and we have
Next we consider (
23) and we have
Again by Definition 5, we can combine the (
24) and (
25)
Hence the proof. □
Remark 5. Proposition 3, states that --superquadratic is --superquadratic as well.
Proposition 4. Let be the --superquadratic such that and , then is --superquadratic .
Proof. Since for every
as well as
, one derives that
Again by Definition 5, the result (
26) can be split as
and
First consider (
27), we get
Similarly by considering (
28), we get
Again using Definition 5, the results (
29) and (
30) can be combined as follows:
Consequently, is termed as a --superquadratic . □
Theorem 6. Let be a --superquadratic for all points such that such that , then the inequalityholds, . Proof. Let
is a
-
-superquadratic
, then
By Definition 5, the result (
31) can be split as follows:
and
Next putting
and
in (
34), we get
Substituting
,
and
in (
32), we get
Similarly by considering (
33) and moving in the same fashion as before, we get
Again by Definition 5, the inequalities (
36) and (
37) can be merged and we get the required result. □
Theorem 7 (Jensen inequality). Let , and . If ψ is a --superquadratic on such that , thenwhere . Proof. Since by Definition 5, the result (
38) can be split as follows:
and
First considering (
39) and (
40) and setting
, we get
and
respectively. Again by Definition 5, the results (
41) and (
42) can be combined as follows:
Thus, the statement (
38) is true for
.
Let us assume that the relation (
38) holds true for
. Consequently, we get
Now, let us demonstrate that (
38) holds for
, so we first consider the center value of the function
where
, then it implies that
Using the center value result of (
44) in (
45), we obtain
Likewise, taking the radius values into account and following the same steps, we attain
Again by Definition 5, the results (
46) and (
47) can be combined as follows:
Hence the proof. □
Next, we provide the proof of -type inequalities for --superquadratic .
Theorem 8. Let be a --superquadratic such that , for . If with , thenholds. Proof. Since
is a
-
-superquadratic
, therefore we have
By Definition 5, the result (
49) can be split as follows:
and
First considering (
50) and changing
by
and
by
, we attain
Integrating (
52) with respect to
over
, we have
Similarly considering (
51), and moving in the same way, we get
Again by Definition 5, the results (
53) and (
54) can be merged as follows:
As
is a
-superquadratic, it follows that
Integrating (
54) with respect to
over
, we have
Similarly considering
as
-superquadratic function, we get
Again by Definition 5, the results (
58) and (
59) can be merged as follows:
Merging (
55) and (
60), we get the required result. □
Corollary 1. Reversing the inequality in (48) yields the -inequalities for --subquadratic s. Corollary 2. If is put in (48), we attain -inequalities for -P-superquadratic function: The following example describes the truth of the statement of Theorem 8 for the --superquadratic function.
Example 2. Since is a --superquadratic function on for as shown in Example 1, then Figure 2 and Table 1, describe the right, middle and left terms denoted by , and , respectively, for the Theorem 8. Similarly considering the function
, we get following
Figure 3 and
Table 2 showing the authenticity of Theorem 8.
4. Applications to Information Theory
The task of distinguishing between two probability distributions is fundamental in fields such as statistics, information theory, and machine learning. Divergence measures serve as quantitative instruments for evaluating the dissimilarity between probability distributions. In 1991, Lin [
27] introduced a new class of divergence measures derived from Shannon entropy, a core concept in information theory employed to quantify both uncertainty and information content within a probability distribution. Lin’s divergence offered a principled, information theoretic approach to measuring distributional discrepancy. Building upon this work, Shioya and Da-te [
28] extended Lin’s method in 1995 by introducing the
-divergence. Their formulation leveraged the
inequality, a classical mathematical result associated with convex functions, to establish a more general and flexible framework. As a result of this advancement, the applicability of divergence measures was greatly extended, facilitating novel analyses and insights into probability distributions.
In order to ensure clarity, we limit ourselves to the definitions that are relevant to the results obtained below.
Definition 9 (Csiszár -divergence [29]). Here, represents the class of all probability densities with respect to a σ-finite measure δ, , and ψ is convex on . Remark 6. Taking ψ to be a -superquadratic function yields the Csiszár ψ-divergence for -superquadratic functions.
Definition 10 ( -divergence [30]). Definition 11 ( fractional -divergence [31]).where Theorem 3, specifies the fractional integrals utilized in (76). Remark 7. By taking in (76), we recover (75). We now establish the results associated with fractional -divergence and -divergence for -(P, m)-superquadratic :
Theorem 10. Let , on , is a -(P, m)-superquadratic and , then
∀
, whereand Proof. The
-type inequalities for
-(P, m)-superquadratic
, as stated in Theorem 8, are considered.
Now setting
, and
, in (
78), we obtain
Since
, it follows that
For all
, multiply both sides of (
80) by
and integrate over ℧ to obtain
Hence, the desired conclusion follows from the definitions of divergence and -divergence. □
Theorem 11. Let , on , is a -(P, m)-superquadratic and , then ∀
, and , whereand Proof. Theorem 9 presents fractional
-type inequalities for
-(P, m)-superquadratic
formulated using
integral operators of order
.
Now setting
, and
, in (
82), we obtain
Given that
, it implies that
Multiplying (
83), both sides by
, where
and then integrating the result on ℧, we obtain:
Employing the definitions of divergence, -divergence and fractional -divergence, consequently, the stated result is established. □
Remark 8. The result of Theorem 10 follows directly by taking in Theorem 11.
5. Conclusions
In this study, we introduced and systematically developed a novel class of --superquadratic and investigated their fundamental structural properties. By employing these properties, we successfully derived new forms of Jensen and -type inequalities, along with their fractional counterparts involving the fractional integral operators within the framework of interval calculus. The analytical results were further validated through illustrative numerical computations and graphical demonstrations, confirming the robustness and generality of the proposed approach. Additionally, applications to information theory highlight the theoretical significance and practical relevance of the presented findings. The developed results not only generalize but also refine several existing inequalities reported in the literature, thereby enriching the landscape of modern inequality theory and fractional analysis.
Future research may proceed in several promising directions.
Problem 1. Fractional Inequalities of --superquadratic via Generalized Proportional Fractional Operators.
Problem 2. Fractional Perspective of --superquadratic via Katugampola Fractional Operators.
Problem 3. Fractional Variants of --superquadratic via Atangana-Baleanu Fractional Operators.
Problem 4. Fractional Estimates of --superquadratic via Conformable Fractional Operators.
Another potential avenue is to investigate stochastic, fuzzy, or quantum variants of these inequalities, which may uncover deeper insights into uncertainty modeling and quantum information theory. Moreover, extending the established results to multi-dimensional or operator-valued settings could open up further applications in functional analysis, optimization, and applied mathematical modeling.