INFORMATION MENU
Full-Text PDF

Engineering and Applied Science Letter

Vol. 3 (2020), Issue 2, pp. 9 – 18
ISSN: 2617-9709 (Online) 2617-9695 (Print)
DOI: 10.30538/psrp-easl2020.0037

New fractional Hadamard and Fejér-Hadamard inequalities associated with exponentially \((h,m)\)-convex functions

Sajid Mehmood, Ghulam Farid\(^1\), Khuram Ali Khan, Muhammad Yussouf
Department of Mathematics, COMSATS University Islamabad, Attock Campus, Pakistan.; (S.M & G.F)
Department of Mathematics, University of Sargodha, Sargodha, Pakistan.; (K.A.K & M.Y)

\(^{1}\)Corresponding Author: ghlmfarid@cuiatk.edu.pk
Copyright ©2020 Sajid Mehmood, Ghulam Farid, Khuram Ali Khan, Muhammad Yussouf. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Received: March 1, 2020 – Accepted: April 5, 2020 – Published: April 9, 2020

Abstract

The aim of this paper is to establish some new fractional Hadamard and Fejér-Hadamard inequalities for exponentially \((h,m)\)-convex functions. These inequalities are produced by using the generalized fractional integral operators containing Mittag-Leffler function via a monotonically increasing function. The presented results hold for various kinds of convexities and well known fractional integral operators.

Keywords:

Convex functions, exponentially \((h,m)\)-convex functions, Hadamard inequality, Fejér-Hadamard inequality, generalized fractional integral operators, Mittag-Leffler function.

1. Introduction

Convex functions are very important in the field of mathematical inequalities. Nobody can deny the importance of convex functions. A large number of mathematical inequalities exist in literature due to convex functions. For more information related to convex functions and it's properties (see, [1, 2, 3]).

Definition 1. A function \(\mu:I\rightarrow \mathbb{R}\) on an interval of real line is said to be convex, if for all \(\alpha,\beta\in I\) and \(\kappa\in[0,1]\), the following inequality holds:

\begin{equation}\label{34} \mu(\kappa \alpha+(1-\kappa)\beta)\leq \kappa\mu(\alpha)+(1-\kappa)\mu(\beta). \end{equation}
(1)
The function \(\mu\) is said to be concave if \(-\mu\) is convex.

A convex function is interpreted very nicely in the coordinate plane by the well known Hadamard inequality stated as follows:

Theorem 2. Let \(\mu:[\alpha,\beta]\rightarrow \mathbb{R}\) be a convex function such that \(\alpha< \beta\). The following inequalities holds: \begin{equation*} \mu\left(\frac{\alpha+\beta}{2}\right)\leq \frac{1}{\beta-\alpha}\int^{\beta}_{\alpha}\mu(\kappa)d\kappa\leq \frac{\mu(\alpha)+\mu(\beta)}{2}. \end{equation*}

In [4], Fejér gave the generalization of Hadamard inequality known as the Fejér-Hadamard inequality stated as follows:

Theorem 3. Let \(\mu:[\alpha,\beta]\rightarrow \mathbb{R}\) be a convex function such that \(\alpha< \beta\). Also let \(\nu:[\alpha,\beta]\rightarrow \mathbb{R}\) be a positive, integrable and symmetric to \(\frac{\alpha+\beta}{2}\). The following inequalities hold:

\begin{equation}\label{r} \mu\left(\frac{\alpha+\beta}{2}\right)\int_{\alpha}^{\beta}\nu(\kappa)d\kappa\leq \int_{\alpha}^{\beta}\mu(\kappa)\nu(\kappa)d\kappa\leq \frac{\mu(\alpha)+\mu(\beta)}{2}\int_{\alpha}^{\beta}\nu(\kappa)d\kappa. \end{equation}
(2)
The Hadamard and the Fejér-Hadamard inequalities are further generalized in various ways by using different fractional integral operators such as Riemann-Liouville, Katugampola, conformable and generalized fractional integral operators containing Mittag-Leffler function etc. For more results and details (see, [5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21]). Next we give the definition of exponentially convex functions.

Definition 4.[9, 22] A function \(\mu: I\rightarrow\mathbb{R}\) on an interval of real line is said to be exponentially convex, if for all \(\alpha,\beta \in I\) and \(\kappa\in[0,1]\), the following inequality holds:

\begin{equation}\label{11} e^{ \mu(\kappa \alpha+(1-\kappa)\beta)}\leq \kappa e^{\mu(\alpha)}+(1-\kappa)e^{\mu(\beta)}. \end{equation}
(3)
In [23], Rashid et al., gave the definition of exponentially \(s\)-convex functions.

Definition 5. Let \(s\in[0,1]\). A function \(\mu: I\rightarrow\mathbb{R}\) on an interval of real line is said to be exponentially \(s\)-convex, if for all \(\alpha,\beta \in I\) and \(\kappa\in[0,1]\), the following inequality holds:

\begin{equation}\label{12} e^{ \mu(\kappa \alpha+(1-\kappa)\beta)}\leq \kappa^s e^{\mu(\alpha)}+(1-\kappa)^se^{\mu(\beta)}. \end{equation}
(4)
In [24], Rashid et al., gave the definition of exponentially \(h\)-convex functions.

Definition 6. Let \(J\subseteq\mathbb{R}\) be an interval containing \((0,1)\) and let \(h: J\rightarrow\mathbb{R}\) be a non-negative function. Then a function \(\mu: I\rightarrow\mathbb{R}\) on an interval of real line is said to be exponentially \(h\)-convex, if for all \(\alpha,\beta \in I\) and \(\kappa\in[0,1]\), the following inequality holds:

\begin{equation}\label{13} e^{ \mu(\kappa \alpha+(1-\kappa)\beta)}\leq h(\kappa) e^{\mu(\alpha)}+h(1-\kappa)e^{\mu(\beta)}. \end{equation}
(5)
In [25], Rashid et al., gave the definition of exponentially \(m\)-convex functions.

Definition 7. A function \(\mu: I\rightarrow\mathbb{R}\) on an interval of real line is said to be exponentially \(m\)-convex, if for all \(\alpha,\beta \in I\), \(m\in(0,1]\) and \(\kappa\in[0,1]\), the following inequality holds:

\begin{equation}\label{14} e^{ \mu(\kappa \alpha+m(1-\kappa)\beta)}\leq \kappa e^{\mu(\alpha)}+m(1-\kappa)e^{\mu(\beta)}. \end{equation}
(6)
In [26], Rashid et al., gave the definition of exponentially \((h,m)\)-convex functions.

Definition 8. Let \(J\subseteq\mathbb{R}\) be an interval containing \((0,1)\) and let \(h: J\rightarrow\mathbb{R}\) be a non-negative function. Then a function \(\mu: I\rightarrow\mathbb{R}\) on an interval of real line is said to be exponentially \((h,m)\)-convex, if for all \(\alpha,\beta \in I\), \(m\in(0,1]\) and \(\kappa\in[0,1]\), the following inequality holds:

\begin{equation}\label{15} e^{ \mu(\kappa \alpha+m(1-\kappa)\beta)}\leq h(\kappa) e^{\mu(\alpha)}+mh(1-\kappa)e^{\mu(\beta)}. \end{equation}
(7)

Remark 1.

  1. If we set \(h(\kappa)=\kappa\) and \(m=1\) in (7), then exponentially convex function (3) is obtained.
  2. If we set \(h(\kappa)=\kappa^s\) and \(m=1\) in (7), then exponentially \(s\)-convex function (4) is obtained.
  3. If we set \(m=1\) in (7), then exponentially \(h\)-convex function (5) is obtained.
  4. If we set \(h(\kappa)=\kappa\) in (7), then exponentially \(m\)-convex function (6) is obtained.

Fractional integral operators also play important role in the subject of mathematical analysis. Recently in [27], Andrić et al., defined the generalized fractional integral operators containing generalized Mittag-Leffler function in their kernels as follows:

Definition 9. Let \(\psi,\sigma,\phi,l,\varsigma,c\in \mathbb{C}\), \(\Re(\sigma),\Re(\phi),\Re(l)>0\), \(\Re(c)>\Re(\varsigma)>0\) with \(p\geq0\), \(r>0\) and \(0< q\leq r+\Re(\sigma)\). Let \(\mu\in L_{1}[\alpha,\beta]\) and \(u\in[\alpha,\beta].\) Then the generalized fractional integral operators \(\Upsilon_{\sigma,\phi,l,\psi,\alpha^{+}}^{\varsigma,r,q,c}\mu \) and \(\Upsilon_{\sigma,\phi,l,\psi,\beta^{-}}^{\varsigma,r,q,c}\mu\) are defined by:

\begin{equation}\label{a} \left(\Upsilon_{\sigma,\phi,l,\psi,\alpha^{+}}^{\varsigma,r,q,c}\mu \right)(u;p)=\int_{\alpha}^{u}(u-\kappa)^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi(u-\kappa)^{\sigma};p)\mu(\kappa)d\kappa, \end{equation}
(8)
\begin{equation} \left(\Upsilon_{\sigma,\phi,l,\psi,\beta^{-}}^{\varsigma,r,q,c}\mu \right)(u;p)=\int_{u}^{\beta}(\kappa-u)^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi(\kappa-u)^{\sigma};p)\mu(\kappa)d\kappa, \end{equation}
(9)
where \(E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\kappa;p)\) is the generalized Mittag-Leffler function defined as follows: \begin{equation*} E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\kappa;p)= \sum_{n=0}^{\infty}\frac{\beta_{p}(\varsigma+nq,c-\varsigma)}{\beta(\varsigma,c-\varsigma)} \frac{(c)_{nq}}{\Gamma(\sigma n +\phi)} \frac{\kappa^{n}}{(l)_{n r}}. \end{equation*}

In [28], Farid defined the following unified integral operators:

Definition 10. Let \(\mu, \nu: [\alpha,\beta]\rightarrow \mathbb{R}\), \(0< \alpha< \beta\) be the functions such that \(\mu\) be a positive and integrable and \(\nu\) be a differentiable and strictly increasing. Also, let \(\frac{\gamma}{u}\) be an increasing function on \([\alpha,\infty)\) and \(\psi,\phi,l,\varsigma,c\in \mathbb{C}\), \(\Re(\phi),\Re(l)>0\), \(\Re(c)>\Re(\varsigma)>0\) with \(p\geq0\), \(\sigma,r>0\) and \(0< q\leq r+\sigma\). Then for \(u\in[\alpha,\beta]\) the integral operators \(_{\nu}\Upsilon_{\sigma, \phi,l, \alpha^{+}}^{\gamma, \varsigma,r,q,c}\mu\) and \(_{\nu}\Upsilon_{\sigma, \phi,l, \beta^{-}}^{\gamma, \varsigma,r,q,c}\mu\) are defined by:

\begin{align}\label{sd} \left(_{\nu}\Upsilon_{\sigma, \phi,l, \alpha^{+}}^{\gamma, \varsigma,r,q,c}\mu\right)(u;p)=\int_{\alpha}^{u}\frac{\gamma(\nu(u)-\nu(\kappa))}{\nu(u)-\nu(\kappa)} E_{\sigma, \phi, l}^{\varsigma,r,q,c} (\psi(\nu(u)-\nu(\kappa))^{\sigma};p)\mu(\kappa)d(\nu(\kappa)), \end{align}
(10)
\begin{align}\label{sb} \left(_{\nu}\Upsilon_{\sigma, \phi,l, \beta^{-}}^{\gamma, \varsigma,r,q,c}\mu\right)(u;p)=\int_{u}^{\beta}\frac{\gamma(\nu(\kappa)-\nu(u))}{\nu(\kappa)-\nu(u)} E_{\sigma, \phi, l}^{\varsigma,r,q,c} (\psi(\nu(\kappa)-\nu(u))^{\sigma};p)\mu(\kappa)d(\nu(\kappa)). \end{align}
(11)
If we set \(\gamma(u)=u^\phi\) in (10) and (11), then we get the following generalized fractional integral operators containing Mittag-Leffler function:

Definition 11. Let \(\mu, \nu: [\alpha,\beta]\rightarrow \mathbb{R}\), \(0< \alpha< \beta\) be the functions such that \(\mu\) be a positive and integrable and \(\nu\) be a differentiable and strictly increasing. Also let \(\psi,\phi,l,\varsigma,c\in \mathbb{C}\), \(\Re(\phi),\Re(l)>0\), \(\Re(c)>\Re(\varsigma)>0\) with \(p\geq0\), \(\sigma,r>0\) and \(0< q\leq r+\sigma\). Then for \(u\in[\alpha,\beta]\) the integral operators \(_{\nu}\Upsilon_{\sigma, \phi,l,\psi, \alpha^{+}}^{\varsigma,r,q,c}\mu\) and \(_{\nu}\Upsilon_{\sigma, \phi,l,\psi, \beta^{-}}^{ \varsigma,r,q,c}\mu\) are defined by:

\begin{equation}\label{1} \left(_{\nu}\Upsilon_{\sigma, \phi,l, \psi, \alpha^{+}}^{ \varsigma,r,q,c}\mu\right)(u;p)=\int_{\alpha}^{u}(\nu(u)-\nu(\kappa))^{\phi-1} E_{\sigma, \phi, l}^{\varsigma,r,q,c} (\psi(\nu(u)-\nu(\kappa))^{\sigma};p)\mu(\kappa)d(\nu(\kappa)), \end{equation}
(12)
\begin{equation}\label{7} \left(_{\nu}\Upsilon_{\sigma, \phi,l,\psi, \beta^{-}}^{ \varsigma,r,q,c}\mu\right)(u;p)=\int_{u}^{\beta}(\nu(\kappa)-\nu(u))^{\phi-1} E_{\sigma, \phi, l}^{\varsigma,r,q,c} (\psi(\nu(\kappa)-\nu(u))^{\sigma};p)\mu(\kappa)d(\nu(\kappa)). \end{equation}
(13)

Remark 2. (12) and (13) are the generalization of the following fractional integral operators:

  1. Setting \(\nu(u)=u\), the fractional integral operators (8) and (9), can be obtained.
  2. Setting \(\nu(u)=u\) and \(p=0\), the fractional integral operators defined by Salim-Faraj in [29]TOS}, can be obtained.
  3. Setting \(\nu(u)=u\) and \(l=r=1\), the fractional integral operators defined by Rahman et al., in [30], can be obtained.
  4. Setting \(\nu(u)=u\), \(p=0\) and \(l=r=1\), the fractional integral operators defined by Srivastava-Tomovski in [31], can be obtained.
  5. Setting \(\nu(u)=u\), \(p=0\) and \(l=r=q=1\), the fractional integral operators defined by Prabhakar in [32], can be obtained.
  6. Setting \(\nu(u)=u\) and \(\psi=p=0\), the Riemann-Liouville fractional integral operators can be obtained.

In [33], Mehmood et al., proved the following formulas for constant function:
\begin{equation}\label{1*} \left(_{\nu}\Upsilon_{\sigma, \phi,l, \psi, \alpha^{+}}^{ \varsigma,r,q,c}1\right)(u;p)=(\nu(u)-\nu(\alpha))^{\phi}E_{\sigma, \phi+1, l}^{\varsigma,r,q,c} (\psi(\nu(u)-\nu(\alpha))^{\sigma};p):=_{\nu}\xi_{\psi, \alpha^{+}}^{\phi}(u;p), \end{equation}
(14)
\begin{equation}\label{7*} \left(_{\nu}\Upsilon_{\sigma, \phi,l, \psi, \beta^{-}}^{ \varsigma,r,q,c}1\right)(u;p)=(\nu(\beta)-\nu(u))^{\phi}E_{\sigma, \phi+1, l}^{\varsigma,r,q,c} (\psi(\nu(\beta)-\nu(u))^{\sigma};p):=_{\nu}\xi_{\psi, \beta^{-}}^{ \phi}(u;p). \end{equation}
(15)
The objective of this paper is to establish the Hadamard and the Fejér-Hadamard inequalities for generalized fractional integral operators (12) and (13) containing Mittag-Leffler function via a monotone function by using the exponentially \((h,m)\)-convex functions. These inequalities lead to produce the Hadamard and the Fejér-Hadamard inequalities for various kinds of exponentially convexity and well known fractional integral operators given in Remark 1 and Remark 2. In Section 2, we prove the Hadamard inequalities for generalized fractional integral operators (12) and (13) via exponentially \((h,m)\)-convex functions. In Section 3, we prove the Fejér-Hadamard inequalities for these generalized fractional integral operators via exponentially \((h,m)\)-convex functions. Moreover, some of the results published in [26, 33, 34] have been obtained in particular.

2.Fractional Hadamard inequalities for exponentially \((h,m)\)-convex functions

In this section, we will give two versions of the generalized fractional Hadamard inequality. To establish these inequalities exponentially \((h,m)\)-convexity and generalized fractional integrals operators have been used.

Theorem 12. Let \(\mu,\nu: [\alpha,m\beta]\subset[0,\infty)\to\mathbb{R}\), \(0< \alpha< m\beta\) be two functions such that \(\mu\) be integrable and \(\nu\) be differentiable. If \(\mu\) be exponentially \((h,m)\)-convex, \(\nu\) be strictly increasing and \(h\in[0,1]\). Then for generalized fractional integral operators, the following inequalities hold:

\begin{align}\label{yty} &{e^{\mu\left( \frac{\nu(\alpha)+m\nu(\beta)}{2}\right)}} _{\nu}\xi^{\phi}_{\bar{\psi},\alpha^{+}}(\nu^{-1}(m\nu(\beta));p)\nonumber\\\nonumber &\leq\! h\!\left(\!\frac{1}{2}\!\right)\!\!\left[\!\left(\!_{\nu}\Upsilon_{\sigma,\phi,l,\bar{\psi},\alpha^{+}}^{\varsigma,r,q,c}e^{\mu\circ\nu}\!\right)\!\!(\!\nu^{-1}(\!m\nu(\beta));p)\!+\!\nonumber m^{\phi+1}\!\!\left(\!_{\nu}\Upsilon_{\sigma,\phi,l,\bar{\psi}m^\sigma,\beta^{-}}^{\varsigma,r,q,c}e^{\mu\circ\nu}\!\!\right)\!\!\left(\!\!\nu^{-1}\!\left(\!\frac{\nu(\alpha)}{m}\!\right)\!;p\!\right)\!\right]\\\nonumber &\leq h\left(\frac{1}{2}\right){(m\nu(\beta)-\nu(\alpha))}^\phi\left[\left(e^{\mu(\nu(\alpha))}+me^{\mu(\nu(\beta))}\right) \left(\Upsilon_{\sigma,\phi,l,{\psi},1^{-}}^{\varsigma,r,q,c}h\right)(0;p)\right.\nonumber\\ &\;\;\;\left.+m\left(e^{\mu(\nu(\beta))}+me^{\mu\left(\frac{\nu(\alpha)}{m^2}\right)}\right)\left(\Upsilon_{\sigma,\phi,l,{\psi},0^{+}}^{\varsigma,r,q,c}h\right)(1;p)\right],\;\;\text{where}\;\; \bar{\psi}=\frac{\psi}{(m\nu(\beta)-\nu(\alpha))^{\sigma}}. \end{align}
(16)

Proof. By the exponentially \((h,m)\)-convexity of \(\mu\), we have

\begin{equation}\label{b} {e^{\mu\left( \frac{\nu(\alpha)+m\nu(\beta)}{2}\right)}}\leq h\left(\frac{1}{2}\right) \left[{e^{{\mu(\kappa\nu(\alpha)+m(1-\kappa)\nu(\beta))}}}+m{e^{{\mu((1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}}}\right]. \end{equation}
(17)
Multiplying (17) with \(\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)\) and integrating over \([0,1]\), we have
\begin{align}\label{ad} &{e^{\mu\left( \frac{\nu(\alpha)+m\nu(\beta)}{2}\right)}}\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)d\kappa\notag \\&\leq h\left(\frac{1}{2}\right) \left[ \int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p){e^{{\mu(\kappa\nu(\alpha)+m(1-\kappa)\nu(\beta))}}}d\kappa\right.\left.+m\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p){e^{{\mu((1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}}}d\kappa\right]. \end{align}
(18)
Setting \(\nu(u)=\kappa\nu(\alpha)+m(1-\kappa)\nu(\beta)\) and \(\nu(v)=(1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta)\) in (18), then again from exponentially \((h,m)\)-convexity of \(\mu\), we have
\begin{align}\label{c} &e^{{\mu(\kappa\nu(\alpha)+m(1-\kappa)\nu(\beta))}}+me^{{\mu((1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}}\notag\\ &\leq h(\kappa)\left(e^{\mu(\nu(\alpha))}+me^{\mu(\nu(\beta))}\right)+mh(1-\kappa)\left(e^{\mu(\nu(\beta))}+me^{\mu\left(\frac{\nu(\alpha)}{m^2}\right)}\right). \end{align}
(19)
Multiplying (19) with \(h\left(\frac{1}{2}\right)\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)\) and integrating over \([0,1]\), we have
\begin{align}\label{ab} &\notag h\left(\frac{1}{2}\right)\bigg[\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{{\mu(\kappa\nu(\alpha)+m(1-\kappa)\nu(\beta))}}d\kappa+m\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{{\mu((1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}}d\kappa\bigg]\\&\leq h\left(\frac{1}{2}\right) \bigg[\left(e^{\mu(\nu(\alpha))}+me^{\mu(\nu(\beta))}\right)\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)h(\kappa)d\kappa+m\left(e^{\mu(\nu(\beta))}+me^{\mu\left(\frac{\nu(\alpha)}{m^2}\right)}\right) \nonumber\\&\;\;\;\times \int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)h(1-\kappa)d\kappa\bigg]. \end{align}
(20)
Setting \(\nu(u)=\kappa\nu(\alpha)+m(1-\kappa)\nu(\beta)\) and \(\nu(v)=(1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta)\) in (20), then by using (8), (9), (12)and (13), the second inequality of (16) is obtained.

Corollary 1. Setting \(m=1\) in (16), the following inequalities for exponentially \(h\)-convex function can be obtained:

\begin{align}\label{yty*} & {e^{\mu\left( \frac{\nu(\alpha)+\nu(\beta)}{2}\right)}} _{\nu}\xi^{\phi}_{\bar{\psi},\alpha^{+}}(\beta;p)\leq h\left(\frac{1}{2}\right)\bigg[\left(_{\nu}\Upsilon_{\sigma,\phi,l,\bar{\psi},\alpha^{+}}^{\varsigma,r,q,c}e^{\mu\circ\nu}\right)(\beta;p)+ \left(_{\nu}\Upsilon_{\sigma,\phi,l,\bar{\psi},\beta^{-}}^{\varsigma,r,q,c}e^{\mu\circ\nu}\right)\left(\alpha;p \right)\bigg]\nonumber\\ &\leq h\left(\frac{1}{2}\right){(\nu(\beta)-\nu(\alpha))}^\phi\left(e^{\mu(\nu(\alpha))}+e^{\mu(\nu(\beta))}\right)\left[ \left(\Upsilon_{\sigma,\phi,l,{\psi},1^{-}}^{\varsigma,r,q,c}h\right)(0;p)+\left(\Upsilon_{\sigma,\phi,l,{\psi},0^{+}}^{\varsigma,r,q,c}h\right)(1;p)\right], \end{align}
(21)
where \(\bar{\psi}=\frac{\psi}{(\nu(\beta)-\nu(\alpha))^{\sigma}}. \)

Remark 3.

  1. If we set \(h(\kappa)=\kappa\) in (16), then [33, Theorem 8] is obtained.
  2. If we set \(h(\kappa)=\kappa\) and \(m=1\) in (16), then [33, Corollary 1] is obtained.
  3. If we set \(\nu(u)=u\) and \(h(\kappa)=\kappa\) in (16), then [34, Theorem 2.1] is obtained.
  4. If we set \(\nu(u)=u\), \(h(\kappa)=\kappa\) and \(m=1\) in (16), then [34, Corollary 2.2] is obtained.
  5. If we set \(\nu(u)=u\) in (16), then [26, Theorem 2.1] is obtained.

In the following we give another version of the Hadamard inequality for generalized fractional integral operators via exponentially \((h,m)\)-convex functions.

Theorem 13. Let \(\mu,\nu: [\alpha,m\beta]\subset[0,\infty)\to\mathbb{R}\), \(0< \alpha< m\beta\) be two functions such that \(\mu\) be integrable and \(\nu\) be differentiable. If \(\mu\) be exponentially \((h,m)\)-convex and \(\nu\) be strictly increasing. Then for generalized fractional integral operators, the following inequalities hold:

\begin{align}\label{lmo} &\notag{e^{\mu\left( \frac{\nu(\alpha)+m\nu(\beta)}{2}\right)}}_{\nu}\xi^{\phi}_{\bar{\psi}2^\sigma,\left(\nu^{-1}\left(\frac{\nu(\alpha)+m\nu(\beta)}{2}\right)\right) ^{+}}(\nu^{-1}(m\nu(\beta));p)\\\nonumber &\leq h\!\left(\frac{1}{2}\right)\! \left[\left(_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi}2^\sigma,\left(\nu^{-1}\left(\frac{\nu(\alpha)+m\nu(\beta)}{2}\right)\right) ^{+}}e^{\mu\circ\nu}\right)(\nu^{-1}(m\nu(\beta));p)\right.\\&\nonumber\;\;\;\left.+m^{\phi+1} \left(_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi}(2m)^\sigma,\left(\nu^{-1}\left(\frac{\nu(\alpha)+m\nu(\beta)}{2m}\right)\right) ^{-}}e^{\mu\circ\nu}\right)\left(\nu^{-1}\left(\frac{\nu(\alpha)}{m}\right);p \right)\right] \\&\nonumber\leq\!h\!\left(\!\frac{1}{2}\!\right)\!\! \frac{(m\nu(\beta)-\nu(\alpha))^\phi}{2^{\phi}}\!\!\left[\!\left(\!e^{\mu(\nu(\alpha))}+me^{\mu(\nu(\beta))}\!\right)\!\!\int_{0}^{1}\!\!\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)h\left(\frac{\kappa}{2}\right)\!d\kappa\right.\\&\;\;\;\left.+m\!\left(\!e^{\mu(\nu(\beta))}+me^{\mu\left(\frac{\nu(\alpha)}{m^2}\!\right)}\right)\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)h\left(\frac{2-\kappa}{2}\right)d\kappa\right], \end{align}
(22)
where \(\bar{\psi}\) is same as in (16).

Proof. By the exponentially \((h,m)\)-convexity of \(\mu\), we have

\begin{align}\label{d} {e^{\mu\left( \frac{\nu(\alpha)+m\nu(\beta)}{2}\right)}} \leq h\left(\frac{1}{2}\right)\left[{e^{\mu\left( \frac{\kappa}{2}\nu(\alpha)+m\frac{(2-\kappa)}{2}\nu(\beta)\right)}}+m {e^{\mu\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}}\right]. \end{align}
(23)
Multiplying (23) with \(\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)\) and integrating over \([0,1]\), we have
\begin{align}\label{w} &{e^{\mu\left( \frac{\nu(\alpha)+m\nu(\beta)}{2}\right)}}\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)d\kappa \nonumber\\ &\leq h\left(\frac{1}{2}\right) \bigg[\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p){e^{\mu\left( \frac{\kappa}{2}\nu(\alpha)+m\frac{(2-\kappa)}{2}\nu(\beta)\right)}}d\kappa+m\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\mu\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}}d\kappa\bigg]. \end{align}
(24)
Setting \(\nu(u)=\frac{\kappa}{2}\nu(\alpha)+m\frac{(2-\kappa)}{2}\nu(\beta)\) and \(\nu(v)=\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\) in (24), then by using (12), (13) and (14), the first inequality of (22) is obtained. Again from exponentially \((h,m)\)-convexity of \(\mu\), we have
\begin{align}\label{e} &e^{\mu\left( \frac{\kappa}{2}\nu(\alpha)+m\frac{(2-\kappa)}{2}\nu(\beta)\right)} +me^{\mu\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)} \\\nonumber &\leq h\left(\frac{\kappa}{2}\right)\left(e^{\mu(\nu(\alpha))}+me^{\mu(\nu(\beta))}\right)+mh\left(\frac{2-\kappa}{2}\right)\left(e^{\mu(\nu(\beta))}+me^{\mu\left(\frac{\nu(\alpha)}{m^2}\right)}\right)\nonumber. \end{align}
(25)
Multiplying (25) with \(h\left(\frac{1}{2}\right) \kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)\) and integrating over \([0,1]\), we have
\begin{align}\label{x} &h\left(\frac{1}{2}\right)\bigg[\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\mu\left( \frac{\kappa}{2}\nu(\alpha)+m\frac{(2-\kappa)}{2}\nu(\beta)\right)}d\kappa+m\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\mu\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}d\kappa\nonumber\bigg]\\&\nonumber\leq h\left(\frac{1}{2}\right) \bigg[ \left(e^{\mu(\nu(\alpha))}+me^{\mu(\nu(\beta))}\right)\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)h\left(\frac{\kappa}{2}\right)d\kappa\nonumber\\&\;\;\;+m\left(e^{\mu(\nu(\beta))}+me^{\mu\left(\frac{\nu(\alpha)}{m^2}\right)}\right) \int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)h\left(\frac{2-\kappa}{2}\right)d\kappa\bigg]. \end{align}
(26)
Putting \(\nu(u)=\frac{\kappa}{2}\nu(\alpha)+m\frac{(2-\kappa)}{2}\nu(\beta)\) and \(\nu(v)=\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\) in (26), then by using (12) and (13), the second inequality of (22) is obtained.

Corollary 2. Setting \(m=1\) in (22), the following inequalities for exponentially \(h\)-convex function can be obtained:

\begin{align}\label{lmo*} &\nonumber2{e^{\mu\left( \frac{\nu(\alpha)+\nu(\beta)}{2}\right)}}_{\nu}\xi^{\phi}_{\bar{\psi}2^\sigma,\left(\nu^{-1}\left(\frac{\nu(\alpha)+\nu(\beta)}{2}\right)\right) ^{+}}(\beta;p)\\\nonumber &\leq h\left(\frac{1}{2}\right) \left[\left(_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi}2^\sigma,\left(\nu^{-1}\left(\frac{\nu(\alpha)+\nu(\beta)}{2}\right)\right) ^{+}}e^{\mu\circ\nu}\right)(\beta;p)\right.\left.+ \left(_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi}2^\sigma,\left(\nu^{-1}\left(\frac{\nu(\alpha)+\nu(\beta)}{2}\right)\right) ^{-}}e^{\mu\circ\nu}\right)(\alpha;p)\right]\\ &\leq h\left(\frac{1}{2}\right) \frac{(\nu(\beta)-\nu(\alpha))^\phi}{2^{\phi}}\left(\!e^{\mu(\nu(\alpha))}+e^{\mu(\nu(\beta))}\right)\!\left[\!\int_{0}^{1}\!\!\!\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)h\left(\frac{\kappa}{2}\right)\!d\kappa\!+\!\!\!\int_{0}^{1}\!\!\!\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)h\left(\!\frac{2-\kappa}{2}\!\right)\!d\kappa\right] \end{align}
(27)
where \(\bar{\psi}\) is same as in (21).

Remark 4.

  1. If we set \(h(\kappa)=\kappa\) in (22), then [33, Theorem 9] is obtained.
  2. If we set \(h(\kappa)=\kappa\) and \(m=1\) in (22), then [33, Corollary 2] is obtained.
  3. If we set \(\nu(u)=u\) and \(h(\kappa)=\kappa\) in (22), then [34, Theorem 2.4] is obtained.
  4. If we set \(\nu(u)=u\), \(h(\kappa)=\kappa\) and \(m=1\) in (22), then [34, Corollary 2.5] is obtained.
  5. If we set \(\nu(u)=u\) in (22), then [26, Theorem 2.2] is obtained.

3.Fractional Fejér-Hadamard Inequalities for exponentially \((h,m)\)-convex functions

In this section, we will give two versions of the generalized fractional Fejér-Hadamard inequality. To establish these inequalities exponentially \((h,m)\)-convexity and generalized fractional integrals operators have been used.

Theorem 14. Let \(\mu,\nu: [\alpha,m\beta]\subset[0,\infty)\to\mathbb{R}\), \(0< \alpha< m\beta\) be two functions such that \(\mu\) be integrable and \(\nu\) be differentiable. If \(\mu\) be exponentially \((h,m)\)-convex and \(\mu(\nu(v))=\mu(\nu(\alpha)+m\nu(\beta)-m\nu(v))\) and \(\nu\) be strictly increasing. Also, let \(\gamma : [\alpha,m\beta]\to\mathbb{R}\) be a function which is non-negative and integrable. Then for generalized fractional integral operators, the following inequalities hold:

\begin{align}\label{yy} &\notag {e^{\mu\left( \frac{\nu(\alpha)+m\nu(\beta)}{2}\right)}} \left(_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi}m^\sigma,\beta^{-}}e^{\gamma\circ\nu}\right)\left(\nu^{-1}\left(\frac{\nu(\alpha)}{m}\right);p\right) \leq {h\left(\frac{1}{2}\right)(1+m)\left(_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi}m^\sigma,\beta^{-}}e^{\mu\circ\nu}e^{\gamma\circ\nu}\right)\left(\nu^{-1}\left(\frac{\nu(\alpha)}{m}\right);p\right)}\\\nonumber &\leq h\left(\frac{1}{2}\right) \frac{(m\nu(\beta)-\nu(\alpha))^\phi}{m^\phi}\bigg[\left(e^{\mu(\nu(\alpha))}+me^{\mu(\nu(\beta))}\right)\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}h\left(\kappa\right)d\kappa\\ &\;\;\;+\!m\!\left(\!e^{\mu(\nu(\beta))}\!+\!me^{\mu\left(\frac{\nu(\alpha)}{m^2}\right)}\!\right) \!\!\int_{0}^{1}\!\!\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}h\left(1-\kappa\right)\!d\kappa\bigg],\, \end{align}
(28)
where \(\bar{\psi}\) is same as in (16).

Proof. Multiplying (17) with \(\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}\) and integrating over \([0,1]\), we have

\begin{align}\label{z} &\notag{e^{\mu\left( \frac{\nu(\alpha)+m\nu(\beta)}{2}\right)}}\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))} d\kappa\\\nonumber & \leq h\left(\frac{1}{2}\right) \bigg[\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p){e^{{\mu(\kappa\nu(\alpha)+m(1-\kappa)\nu(\beta))}}}e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))} d\kappa\\ &\;\;\;+m\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p){e^{{\mu((1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}}}e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}d\kappa\bigg]. \end{align}
(29)
Setting \(\nu(v)=(1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta)\) in (29), then by using (13) and assumption \(\mu(\nu(v))=\mu(\nu(\alpha)+m\nu(\beta)-m\nu(v))\), the first inequality of (28) is obtained. Now multiplying (19) with \(h\left(\frac{1}{2}\right)\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}\) and integrating over \([0,1]\), we have
\begin{align}\label{23} &\notag h\left(\frac{1}{2}\right)\bigg[\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{{\mu(\kappa\nu(\alpha)+m(1-\kappa)\nu(\beta))}}e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))} d\kappa\\ \nonumber&\;\;\;+m\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{{\mu((1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}}e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}d\kappa\bigg]\\&\nonumber\leq\! h\!\left(\!\frac{1}{2}\!\right) \!\!\bigg[\!\!\left(\!e^{\mu(\nu(\alpha))}+me^{\mu(\nu(\beta))}\!\right)\!\!\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}h\left(\kappa\right)d\kappa\\ &\;\;\;+\!m\!\left(\!e^{\mu(\nu(\beta))}\!+\!me^{\mu\left(\frac{\nu(\alpha)}{m^2}\right)}\!\right) \!\!\int_{0}^{1}\!\!\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\gamma( (1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta))}h\left(1-\kappa\right)\!d\kappa\bigg]. \end{align}
(30)
Setting \(\nu(v)=(1-\kappa)\frac{\nu(\alpha)}{m}+\kappa\nu(\beta)\) in (30), then by using (13) and assumption \(\mu(\nu(v))=\mu(\nu(\alpha)+m\nu(\beta)-m\nu(v))\), the second inequality of (28) is obtained.

Corollary 3. Setting \(m=1\) in (28), the following inequalities for exponentially \(h\)-convex function can be obtained:

\begin{align}\label{yy**} &\notag{e^{\mu\left( \frac{\nu(\alpha)+\nu(\beta)}{2}\right)}} \left(_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi},\beta^{-}}e^{\gamma\circ\nu}\right)\left(\alpha;p\right) \leq {2 h\left(\frac{1}{2}\right)\left(_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi},\beta^{-}}e^{\mu\circ\nu}e^{\gamma\circ\nu}\right)\left(\alpha;p\right)}\\\nonumber &\leq h\left(\frac{1}{2}\right) {(\nu(\beta)-\nu(\alpha))^\phi}\left(e^{\mu(\nu(\alpha))}+e^{\mu(\nu(\beta))}\right)\bigg[\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\gamma( (1-\kappa){\nu(\alpha)}+\kappa\nu(\beta))}h\left(\kappa\right)d\kappa\\ &\;\;\;+ \int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\gamma( (1-\kappa){\nu(\alpha)}+\kappa\nu(\beta))}h\left(1-\kappa\right)d\kappa\bigg], \end{align}
(31)
where \(\bar{\psi}\) is same as in (21).

Remark 5.

  1. If we set \(h(\kappa)=\kappa\) in (28), then [33, Theorem 10] is obtained.
  2. If we set \(h(\kappa)=\kappa\) and \(m=1\) in (28), then [33, Corollary 3] is obtained.
  3. If we set \(\nu(u)=u\) and \(h(\kappa)=\kappa\) in (28), then [34, Theorem 2.7] is obtained.
  4. If we set \(\nu(u)=u\), \(h(\kappa)=\kappa\) and \(m=1\) in (28), then [34, Corollary 2.8] is obtained.
  5. If we set \(\nu(u)=u\) in (28), then [26, Theorem 2.3] is obtained.

In the following we give another generalized fractional version of the Fejér-Hadamard inequality.

Theorem 15. Let \(\mu,\nu: [\alpha,m\beta]\subset[0,\infty)\to\mathbb{R}\), \(0< \alpha< m\beta\) be two functions such that \(\mu\) be integrable and \(\nu\) be differentiable. If \(\mu\) be exponentially \((h,m)\)-convex and \(\mu(\nu(v))=\mu(\nu(\alpha)+m\nu(\beta)-m\nu(v))\) and \(\nu\) be strictly increasing. Also, let \(\gamma : [\alpha,m\beta]\to\mathbb{R}\) be a function which is non-negative and integrable. Then for generalized fractional integral operators, the following inequalities hold:

\begin{align}\label{yy*} &\notag{e^{\mu\left( \frac{\nu(\alpha)+m\nu(\beta)}{2}\right)}} \!\!\left(\!_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi}(2m)^\sigma,\left(\nu^{-1}\left(\frac{\nu(\alpha)+m\nu(\beta)}{2m}\right)\right) ^{-}}e^{\gamma\circ\nu}\!\right)\!\left(\!\nu^{-1}\!\left(\frac{\nu(\alpha)}{m}\!\right);p\!\right) \\&\nonumber\leq \!h\!\left(\!\frac{1}{2}\!\right) {\!(1\!+\!m)\!\!\left(\!\!_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi}(2m)^\sigma,\left(\nu^{-1}\left(\frac{\nu(\alpha)+m\nu(\beta)}{2m}\right)\right) ^{-}}e^{\mu\circ\nu}e^{\gamma\circ\nu}\!\!\right)\!\!\left(\!\!\nu^{-1}\!\left(\!\frac{\nu(\alpha)}{m}\right);p\!\right)}\\\nonumber &\leq h\left(\frac{1}{2}\right) \frac{(m\nu(\beta)-\nu(\alpha))^\phi}{(2m)^\phi}\bigg[\left(e^{\mu(\nu(\alpha))}+me^{\mu(\nu(\beta))}\right)\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}}h\left(\frac{\kappa}{2}\right)d\kappa\\ &\;\;\;+\!m\!\!\left(\!e^{\mu(\nu(\beta))}\!+\!me^{\mu\left(\frac{\nu(\alpha)}{m^2}\!\right)}\right)\!\!\!\int_{0}^{1}\!\!\!\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\gamma\left(\!\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\!\right)}}h\!\left(\!\frac{2\!-\!\kappa}{2}\!\right)\!d\kappa\bigg], \end{align}
(32)
where \(\bar{\psi}\) is same as in (16).

Proof. Multiplying (23) with \(\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}}\) and integrating over \([0,1]\), we have

\begin{align}\label{z*} &\notag{e^{\mu\left( \frac{\nu(\alpha)+m\nu(\beta)}{2}\right)}}\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}} d\kappa\\\nonumber & \leq h\left(\frac{1}{2}\right) \bigg[\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p){e^{\mu\left( \frac{\kappa}{2}\nu(\alpha)+m\frac{(2-\kappa)}{2}\nu(\beta)\right)}} {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}} d\kappa\\ &\;\;\;+m\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\mu\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}} {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}}d\kappa\bigg]. \end{align}
(33)
Setting \(\nu(v)=\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\) in 33), then by using (13) and assumption \(\mu(\nu(v))=\mu(\nu(\alpha)+m\nu(\beta)-m\nu(v))\), the first inequality of (32) is obtained. Now multiplying (25) with \(h\left(\frac{1}{2}\right)\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}}\) and integrating over \([0,1]\), we have
\begin{align}\label{24} &\notag h\left(\frac{1}{2}\right)\bigg[\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\mu\left( \frac{\kappa}{2}\nu(\alpha)+m\frac{(2-\kappa)}{2}\nu(\beta)\right)} {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}} d\kappa\\ \nonumber&+m\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p)e^{\mu\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)} {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}}d\kappa\bigg]\\&\nonumber\leq\! h\!\left(\!\frac{1}{2}\!\right)\!\! \bigg[\!\!\left(\!e^{\mu(\nu(\alpha))}\!+\!me^{\mu(\nu(\beta))}\!\right)\!\!\int_{0}^{1}\!\!\!\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\right)}}h\left(\frac{\kappa}{2}\right)\!d\kappa\\ &\;\;\;+\!m\!\!\left(\!e^{\mu(\nu(\beta))}\!+\!me^{\mu\left(\frac{\nu(\alpha)}{m^2}\!\right)}\right)\!\!\!\int_{0}^{1}\!\!\!\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\gamma\left(\!\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\!\right)}}h\!\left(\!\frac{2\!-\!\kappa}{2}\!\right)\!d\kappa\bigg]. \end{align}
(34)
Setting \(\nu(v)=\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}\frac{\nu(\alpha)}{m}\) in (34), then by using (13) and assumption \(\mu(\nu(v))=\mu(\nu(\alpha)+m\nu(\beta)-m\nu(v))\), the second inequality of (32) is obtained.

Corollary 4. Setting \(m=1\) in (32), the following inequalities for exponentially \(h\)-convex function can be obtained:

\begin{align}\label{yy***} &\notag{e^{\mu\left( \frac{\nu(\alpha)+\nu(\beta)}{2}\right)}} \left(_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi}2^\sigma,\left(\nu^{-1}\left(\frac{\nu(\alpha)+\nu(\beta)}{2}\right)\right) ^{-}}e^{\gamma\circ\nu}\right)\left(\alpha;p\right)\leq 2h\left(\frac{1}{2}\right) {\left(_{\nu}\Upsilon^{\varsigma,r,q,c}_{\sigma,\phi,l,\bar{\psi}2^\sigma,\left(\nu^{-1}\left(\frac{\nu(\alpha)+\nu(\beta)}{2}\right)\right) ^{-}}e^{\mu\circ\nu}e^{\gamma\circ\nu}\right)\left(\alpha;p\right)}\\\nonumber &\leq h\left(\frac{1}{2}\right) \frac{(\nu(\beta)-\nu(\alpha))^\phi}{2^\phi}\left(e^{\mu(\nu(\alpha))}+e^{\mu(\nu(\beta))}\right)\bigg[\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}{\nu(\alpha)}\right)}}h\left(\frac{\kappa}{2}\right)d\kappa\\ &\;\;\;+\int_{0}^{1}\kappa^{\phi-1}E_{\sigma,\phi,l}^{\varsigma,r,q,c}(\psi \kappa^{\sigma};p) {e^{\gamma\left(\frac{\kappa}{2}\nu(\beta)+\frac{(2-\kappa)}{2}{\nu(\alpha)}\right)}}h\left(\frac{2-\kappa}{2}\right)d\kappa\bigg], \end{align}
(35)
where \(\bar{\psi}\) is same as in (21).

Remark 6.

  1. If we set \(h(\kappa)=\kappa\) in (32), then [33, Theorem 11] is obtained.
  2. If we set \(h(\kappa)=\kappa\) and \(m=1\) in (32), then [33, Corollary 4] is obtained.

Remark 7. By setting \(h(\kappa)=\kappa^s\) and \(m=1\) in Theorems 12, 13, 14 and 15, the Hadamard and the Fejér-Hadamard inequalities for exponentially \(s\)-convex functions can be obtained. We leave it for interested reader.

4.Concluding remarks

In this article, we established the Hadamard and the Fejér-Hadamard inequalities. To established these inequalities generalized fractional integral operators and exponentially \((h,m)\)-convexity have been used. The presented results hold for various kind of exponentially convexity and well known fractional integral operators given in Remarks 1 and 2. Moreover, the established results have connection with already published results.

Acknowledgments

The research work of the Ghulam Farid is supported by Higher Education Commission of Pakistan under NRPU 2016, Project No. 5421.

Autho Contributions

All authors contributed equally to the writing of this paper. All authors read and approved the final manuscript.

Conflict of Interests

The authors declare no conflict of interest.

References

  1. Niculescu, C., & Persson, L. E. (2006). Convex functions and their applications. A comtemporary approach, CMS Books in Mathematics, vol. 23, Springer Verlag, New York.[Google Scholor]
  2. Pečarić, J., Proschan, F., & Tong, Y. L. (1973). Convex Functions, Partial Orderings and Statistical Applications. Academics Press, New York.[Google Scholor]
  3. Roberts, A. W., & Varberg, D. E. (1973). Convex Functions, Academics Press, New York, USA. [Google Scholor]
  4. Fejér, L. (1906). Überdie Fourierreihen II, Math Naturwiss Anz Ungar Akad Wiss, 24, 369-390.[Google Scholor]
  5. Abbas, G., & Farid, G. (2017). Hadamard and Fejér--Hadamard type inequalities for harmonically convex functions via generalized fractional integrals. The Journal of Analysis, 25(1), 107-119.[Google Scholor]
  6. Awan, M. U., Noor, M. A., & Noor, K. I. (2018). Hermite--Hadamard inequalities for exponentially convex functions. Applied Mathematics & Information Sciences, 12(2), 405-409.[Google Scholor]
  7. Chen, F. (2014). On Hermite-Hadamard type inequalities for Riemann-Liouville fractional integrals via two kinds of convexity. Chinese Journal of Mathematics, 2014, 7 pp.[Google Scholor]
  8. Chen, H., & Katugampola, U. N. (2017). Hermite--Hadamard-Fejér type inequalities for generalized fractional integrals. Journal of Mathematical Analysis and Applications, 446, 1274-1291.[Google Scholor]
  9. Dragomir, S. S., & Gomm, I. (2015). Some Hermite-Hadamard type inequalities for functions whose exponentials are convex. Studia Universitatis Babeș--Bolyai Mathematica, 60(4), 527-534.[Google Scholor]
  10. Farid, G. (2016). Hadamard and Fejér-Hadamard inequalities for generalized fractional integrals involving special functions. Konuralp Journal of Mathematics, 4(1), 108-113.[Google Scholor]
  11. Farid, G. (2018). A treatment of the Hadamard inequality due to m-convexity via generalized fractional integrals. Journal of Fractional Calculus and Applications, 9(1), 8-14.[Google Scholor]
  12. Farid, G., & Abbas, G. (2018). Generalizations of some fractional integral inequalities for m--convex functions via generalized Mittag--Leffler function. Studia Universitatis Babeş--Bolyai Mathematica, 63(1), 23-35.[Google Scholor]
  13. Farid, G., Rehman, A. U., & Tariq, B. (2017). On Hadamard--type inequalities for m--convex functions via Riemann--Liouville fractional integrals. Studia Universitatis Babeş--Bolyai Mathematica, 62(2), 141-150.[Google Scholor]
  14. Farid, G., Rehman, A. U., & Mehmood, S. (2018). Hadamard and Fejér--Hadamard type integral inequalities for harmonically convex functions via an extended generalized Mittag--Leffler function. Journal of Mathematics and Computer Science, 8(5), 630-643.[Google Scholor]
  15. Farid, G., Khan, K. A., Latif, N., Rehman, A. U., & Mehmood, S. (2018). General fractional integral inequalities for convex and m--convex functions via an extended generalized Mittag--Leffler function. Journal of inequalities and applications, 2018, 243 pp.[Google Scholor]
  16. İşcan, İ. (2015). Hermite Hadamard Fejér type inequalities for convex functions via fractional integrals. Studia Universitatis Babeș--Bolyai Mathematica, 60(3), 355-366.[Google Scholor]
  17. Kang, S. M., Farid, G., Nazeer, W., & Mehmood, S. (2019). \((h,m)\)-convex functions and associated fractional Hadamard and Fejér-Hadamard inequalities via an extended generalized Mittag-Leffler function. Journal of inequalities and applications, 2019, 78 pp.[Google Scholor]
  18. Kang, S. M., Farid, G., Nazeer, W., & Tariq, B. (2018). Hadamard and Fejér-Hadamard inequalities for extended generalized fractional integrals involving special functions. Journal of inequalities and applications, 2018, 119 pp.[Google Scholor]
  19. Mehreen, N., & Anwar, M. (2019). Hermite-Hadamard type inequalities for exponentially \(p\)-convex functions and exponentially \(s\)-convex functions in the second sense with applications. Journal of inequalities and applications, 2019, 92 pp.[Google Scholor]
  20. Sarikaya, M. Z., Set, E., Yaldiz, H., & Basak, N. (2013). Hermite--Hadamard's inequalities for fractional integrals and related fractional inequalities. Mathematical and Computer Modelling, 57(9-10), 2403-2407.[Google Scholor]
  21. Sarikaya, M. Z., & Yildirim, H. (2016). On Hermite-Hadamard type inequalities for Riemann-Liouville fractional integrals. Miskolc Mathematical Notes, 17(2), 1049-1059.[Google Scholor]
  22. Antczak, T. (2001). \((p,r)\)-invex sets and functions. Journal of Mathematical Analysis and Applications, 263, 355-379.[Google Scholor]
  23. Rashid, S., Noor, M. A., Noor, K. I., & Akdemir, A. O. (2019). Some new generalizations for exponentially s--convex functions and inequalities via fractional operators. Fractal and Fractional, 3(2), 24.[Google Scholor]
  24. Rashid, S., Noor, M. A., & Noor, K. I. (2019). Some generalize Riemann--Liouville fractional estimates involving functions having exponentially convexity property. Punjab University Journal of Mathematics, 51, 1-15.[Google Scholor]
  25. ras} Rashid, S., Noor, M. A., & Noor, K. I. (2019). Fractional exponentially m-convex functions and inequalities. International Journal of Analysis and Applications, 17(3), 464-478.[Google Scholor]
  26. Rashid, S., Noor, M. A., & Noor, K. I. (2019). Some new estimates for exponentially \((\hbar,\mathfrak {m}) \)-convex functions via extended generalized fractional integral operators. The Korean Journal of Mathematics, 27(4), 843-860.[Google Scholor]
  27. Andrić, M., Farid, G., & Pečarić, J. (2018). A further extension of Mittag-Leffler function. Fractional Calculus and Applied Analysis, 21(5), 1377-1395.[Google Scholor]
  28. Farid, G. (2020). A unified integral operator and further its consequences. Open Journal of Mathematical Analysis, 4(1) (2020), 1-7.[Google Scholor]
  29. Salim, T. O., & Faraj, A. W. (2012). A generalization of Mittag--Leffler function and integral operator associated with fractional calculus. Journal of Fractional Calculus and Applications, 3(5), 1-13.[Google Scholor]
  30. Rahman, G., Baleanu, D., Qurashi, M. A., Purohit, S. D., Mubeen, S., & Arshad, M. (2017). The extended Mittag-Leffler function via fractional calculus. Journal of Nonlinear Sciences and Applications, 10(8), 4244-4253.[Google Scholor]
  31. Srivastava, H. M., & Tomovski, Ž. (2009). Fractional calculus with an integral operator containing a generalized Mittag--Leffler function in the kernel. Applied Mathematics and Computation, 211(1), 198-210.[Google Scholor]
  32. Prabhakar, T. R. (1971). A singular integral equation with a generalized Mittag-Leffler function in the kernel. Yokohama Mathematical Journal, 19, 7-15.[Google Scholor]
  33. Mehmood, S., Farid, G., Khan, K. A., & Yussouf, M. (2020). New Hadamard and Fejér-Hadamard fractional inequalities for exponentially \(m\)-convex function. Engineering and Appllied Sciencie Letters, 3(1), 45-55.[Google Scholor]
  34. Mehmood, S.,& Farid, G. (2020). Fractional Hadamard and Fejér-Hadamard inequalities for exponentially \(m\)-convex function. Studia Universitatis Babeș--Bolyai Mathematica, to appear.[Google Scholor]
Category: EASL - Vol 3 - Issue 2 (2020)