Research Article A Generalized Hermite-Hadamard Inequality ...
Transcript of Research Article A Generalized Hermite-Hadamard Inequality ...
Research ArticleA Generalized Hermite-Hadamard Inequality for CoordinatedConvex Function and Some Associated Mappings
Atiq Ur Rehman,1 Gulam Farid,1 and Sidra Malik2
1COMSATS Institute of Information Technology, Attock Campus, Attock, Pakistan2Government Islamia High School, Attock, Pakistan
Correspondence should be addressed to Atiq Ur Rehman; [email protected]
Received 25 July 2016; Accepted 25 October 2016
Academic Editor: Nan-Jing Huang
Copyright Β© 2016 Atiq Ur Rehman et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.
We have discussed the generalization of Hermite-Hadamard inequality introduced by Lupas for convex functions on coordinatesdefined in a rectangle from the plane. Also we define that mappings are related to it and their properties are discussed.
1. Introduction
Convex functions have great importance in many areas ofMathematics. A real valued function π : πΌ β R, where πΌ isinterval inR, is said to be convex if for πΌ, π½ β₯ 0with πΌ+π½ = 1
π (πΌπ₯ + π½π¦) β€ πΌπ (π₯) + π½π (π¦) . (1)
In particular, if πΌ = π½ = 1/2 we have thatπ(π₯ + π¦2 ) β€ π (π₯) + π (π¦)2 . (2)
One of the most important inequalities that has attractedmany mathematician in this field in the last few decades isthe famous Hermite-Hadamard inequality, which establishesa refinement of (2) and it involves the notions of convexity.
Let π : πΌ β R be a convex mapping defined on theinterval πΌ of real numbers and π, π β πΌwith π < π. The doubleinequality
π(π + π2 ) β€ 1π β π β«
π
ππ (π₯) ππ₯ β€ π (π) + π (π)2 (3)
is known as the Hermite-Hadamard inequality. This inequal-ity was published by Hermite in 1883 and was independentlyproved byHadamard in 1893 (seeMitrinovic and Lackovic [1]for the whole history). It gives us an estimation of the mean
value of a convex function π and it is important to note that(3) provides a refinement to the Jensen inequality. Fink [2] hasworked on a best possible Hermite-Hadamard inequality andhas deduced this inequality with least restrictions. Dragomirin [3, 4] worked on Hermite-Hadamard inequality and amapping in connection to it. Dragomir and Milosevic [5]gave some refinements of Hadamard's inequalities and itsapplications (see also [6β8] for more results). Dragomir andPearce [9] wrote amonograph on selected topics onHermite-Hadamard inequalities.
The aim of this paper is to discuss an analogue of thegeneralization of Hermite-Hadamard inequality introducedby Lupas for convex functions defined in a rectangle fromthe plane. Also some related mappings are defined and theirproperties are discussed.
In [10] Dragomir has given the concept of convex func-tions on the coordinates in a rectangle from the plane andestablished the Hermite-Hadamard inequality for it.
Definition 1. Let Ξ2 = [π, π] Γ [π, π] β R with π < π andπ < π. A function π : Ξ2 β R is called convex on coordinatesif the partial mappings ππ¦ : [π, π] β R, ππ¦(π’) fl π(π’, π¦) andππ₯ : [π, π] β R, ππ₯(V) fl π(π₯, V) are convex, defined for allπ¦ β [π, π] and π₯ β [π, π].
Note that every convex mapping π : Ξ2 β R is convexon the coordinates but the converse is not generally true.
Hindawi Publishing CorporationJournal of MathematicsVolume 2016, Article ID 1631269, 9 pageshttp://dx.doi.org/10.1155/2016/1631269
2 Journal of Mathematics
Theorem 2. Suppose that π : Ξ2 β R is convex on the coordi-nates on Ξ2. Then we have
π(π + π2 , π + π2 ) β€ 12 [1
π β π β«π
ππ(π₯, π + π2 )ππ₯
+ 1π β π β«
π
ππ(π + π2 , π¦) ππ¦] β€ 1
(π β π) (π β π)β β«ππβ«πππ (π₯, π¦) ππ¦ ππ₯ β€ 14 [
1π β π β«
π
ππ (π₯, π) ππ₯
+ 1π β π β«
π
ππ (π₯, π) ππ₯ + 1
π β π β«π
ππ (π, π¦) ππ¦
+ 1π β π β«
π
ππ (π, π¦) ππ¦]
β€ π (π, π) + π (π, π) + π (π, π) + π (π, π)4 .
(4)
In [10], Dragomir for π : Ξ2 β R, defined a mappingπ» : [0, 1]2 β R as
π»(π‘, π ) = 1(π β π) (π β π) β«
π
πβ«πππ(π‘π₯
+ (1 β π‘) π + π2 , π π¦ + (1 β π ) π + π2 )ππ¦ππ₯(5)
and proved the properties of this mapping in followingtheorem.
Theorem 3. Suppose that π : Ξ2 β R is convex on thecoordinates on Ξ2.
(i) Themappingπ» is convex on the coordinates on [0, 1]2.(ii) We have the bounds
sup(π ,π‘)β[0,1]2
π»(π , π‘)
= 1(π β π) (π β π) β«
π
πβ«πππ (π₯, π¦) ππ¦ ππ₯ = π» (1, 1) ,
inf(π ,π‘)β[0,1]2
π»(π , π‘) = π(π + π2 , π + π2 ) = π» (0, 0) .(6)
(iii) The mapping π» is monotonic nondecreasing on thecoordinates.
Also, in [10] Dragomir gave the following mapping,which is closely connected with Hadamard's inequality, οΏ½οΏ½ :[0, 1]2 β R defined as
οΏ½οΏ½ (π‘, π ) fl 1(π β π)2 (π β π)2
β β«ππβ«ππβ«ππβ«πππ (π‘π₯ + (1 β π‘) π¦, π π§ + (1 β π ) π’) ππ₯ ππ¦ ππ§ ππ’.
(7)
and proved the following properties of this mapping.
Theorem 4. Suppose that π : Ξ2 β R is convex on the coordi-nates on Ξ2.
(i) We have the equalities
οΏ½οΏ½ (π‘ + 12 , π ) = οΏ½οΏ½ (12 β π‘, π )βπ‘ β [0, 12] , π β [0, 1] ,
οΏ½οΏ½ (π‘, π + 12) = οΏ½οΏ½ (π‘, 12 β π )βπ‘ β [0, 1] , π β [0, 12] ,
οΏ½οΏ½ (1 β π‘, π ) = οΏ½οΏ½ (π‘, π ) ,οΏ½οΏ½ (π‘, 1 β π ) = οΏ½οΏ½ (π‘, π )
β (π‘, π ) β Ξ2.
(8)
(ii) οΏ½οΏ½ is convex on the coordinates.(iii) We have the bounds
inf(π‘,π )β[0,1]2
οΏ½οΏ½ (π‘, π ) = οΏ½οΏ½ (12 ,12) =
1(π β π)2 (π β π)2
β β«ππβ«ππβ«ππβ«πππ(π₯ + π¦2 , π§ + π’2 ) ππ₯ ππ¦ππ§ ππ’,
sup(π‘,π )β[0,1]2
οΏ½οΏ½ (π‘, π ) = οΏ½οΏ½ (0, 0) = οΏ½οΏ½ (1, 1) = 1(π β π) (π β π)
β β«ππβ«πππ (π₯, π§) ππ§ ππ₯.
(9)
(iv) The mapping οΏ½οΏ½(β , π ) is monotonic nondecreasing on[0, 1/2) and nondecreasing on [1/2, 1] for all π β [0, 1].A similar property has the mapping οΏ½οΏ½(π‘, β ) for all π‘ β[0, 1].
(v) We have the inequality
οΏ½οΏ½ (π‘, π ) β₯ max {π» (π‘, π ) ,π» (1 β π‘, π ) ,π» (π‘, 1 β π ) ,π» (1 β π‘, 1 β π )} . (10)
A generalized form of Hermite-Hadamard inequality isgiven by Lupas in [11] (see also [12, page 143]).
Theorem5. Let p, q be given positive numbers and π < π.Thenthe inequality,
π(ππ + πππ + π ) β€ 12π¦ β«π΄+π¦
π΄βπ¦π (π‘) ππ‘ β€ ππ (π) + ππ (π)π + π , (11)
holds forπ΄ = (ππ+ππ)/(π+π),π¦ > 0 and all continuous convexfunctions π : [π, π] β R iff π¦ β€ ((π β π)/(π + π))min (π, π).
Journal of Mathematics 3
2. Main Results
The following results comprise generalization of Hermite-Hadamard inequality introduced by Lupas on coordinates ina rectangle from the plane.
Theorem 6. Let Ξ2 = [π1, π1] Γ [π2, π2] β R2 and π1, π2, π1,and π2 be positive real numbers and
π΄ π = ππππ + ππππππ + ππ ,π¦π > 0,
where π¦π β€ ππ β ππππ + ππ min (ππ, ππ) for π = 1, 2.(12)
Also let π : Ξ2 β R be convex on the coordinates on Ξ2.π (π΄1, π΄2) β€ 12 [
12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
π (π 1, π΄2) ππ 1
+ 12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
π (π΄1, π 2) ππ 2] β€ 14π¦1π¦2
β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2
β€ 1(π1 + π1) [
π12π¦1 β«π΄1+π¦1
π΄1βπ¦1
π (π 1, π2) ππ 1
+ π12π¦1 β«π΄1+π¦1
π΄1βπ¦1
π (π 1, π2) ππ 1]
+ 1(π2 + π2) [
π22π¦2 β«π΄2+π¦2
π΄2βπ¦2
π (π1, π 2) ππ 2
+ π22π¦2 β«π΄2+π¦2
π΄2βπ¦2
π (π1, π 2) ππ 2] β€ [π1π2π (π1, π2)+ π1π2π (π1, π2) + π1π2π (π1, π2) + π1π2π (π1, π2)]Γ [ 1
(π1 + π1) +1
(π2 + π2)] .
(13)
Proof. Since π : Ξ2 β R is convex on coordinates, it followsthat the mapping ππ
1
: [π2, π2] β R, ππ 1
(π 2) fl π(π 1, π 2), isconvex on [π2, π2] for all π 1 β [π1, π1], so by inequality (11) onehas
ππ 1
(π΄2) β€ 12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
ππ 1
(π 2) ππ 2
β€ π2ππ 1 (π2) + π2ππ 1 (π2)π2 + π2 ,(14)
that is
π (π 1, π΄2) β€ 12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
π (π 1, π 2) ππ 2
β€ π2π (π 1, π2) + π2π (π 1, π2)π2 + π2 .(15)
Observe that for 0 < π¦1 β€ ((π1 β π1)/(π1 + π1))min (π1, π1),by considering two cases (0 < π1 β€ π1 and 0 < π1 < π1) wecan easily verify that π1 β€ π΄1 βπ¦1 < π΄1 +π¦1 β€ π1, so that π isdefined on [π΄1βπ¦1, π΄1+π¦1]. Integrating the above inequalityon [π΄1 β π¦1, π΄1 + π¦1], we have
12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
π (π 1, π΄2) ππ 1 β€ 14π¦1π¦2
β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2
β€ 1π2 + π2 [
π22π¦1 β«π΄1+π¦1
π΄1βπ¦1
π (π 1, π2) ππ 1
+ π22π¦1 β«π΄1+π¦1
π΄1βπ¦1
π (π 1, π2) ππ 1] .
(16)
By a similar argument applied for themappingππ 2
: [π1, π1] βR, ππ
2
(π 1) fl π(π 1, π 2), we get12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
π (π΄1, π 2) ππ 2 β€ 14π¦1π¦2
β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2
β€ 1π1 + π1 [
π12π¦2 β«π΄2+π¦2
π΄2βπ¦2
π (π1, π 2) ππ 2
+ π12π¦2 β«π΄2+π¦2
π΄2βπ¦2
π (π1, π 2) ππ 2] .
(17)
Summing the inequalities (16) and (17), we get the secondand third inequality in (13). Sinceπ is convex on coordinates,using convexity of π on first coordinate and inequality (11),we have
π (π΄1, π΄2) β€ 12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
π (π 1, π΄2) ππ 1 (18)
and, using convexity of π on second coordinate, we have
π (π΄1, π΄2) β€ 12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
π (π΄1, π 2) ππ 2. (19)
Adding the inequalities (18) and (19)we get the first inequalityin (13). Finally, by the same Hermite-Hadamard inequalityintroduced by Lupas we can also state
12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
π (π 1, π2) ππ 1
β€ π1π (π1, π2) + π1π (π1, π2)π1 + π1 ,(20)
12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
π (π 1, π2) ππ 1
β€ π1π (π1, π2) + π1π (π1, π2)π1 + π1 ,(21)
4 Journal of Mathematics
12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
π (π1, π 2) ππ 2
β€ π2π (π1, π2) + π2π (π1, π2)π2 + π2 ,(22)
12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
π (π1, π 2) ππ 2
β€ π2π (π1, π2) + π2π (π1, π2)π2 + π2 .(23)
Multiply (20) and (21) by π2 and multiply (22) and (23) by π2then on addition we get the last inequality in (13).
3. Some Associated Mappings
In this section we will discuss some mappings associatedwith generalized Hermite-Hadamard inequality introducedby Lupas for convex mappings on coordinates.
Let π1, π2, π1, and π2 be positive real numbers and π΄ π =(ππππ + ππππ)/(ππ + ππ) and π¦π > 0 where π¦π β€ ((ππ β ππ)/(ππ +ππ))min(ππ, ππ) for π = 1, 2. For mapping π : Ξ2 = [π1, π1] Γ[π2, π2] β R, we can define a mappingπ» : [0, 1]2 β R,
π»(π’, V) fl 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π’π 1+ (1 β π’)π΄1, Vπ 2 + (1 β V) π΄2) ππ 1ππ 2,
(24)
where π1 β€ π΄1 β π¦1 < π΄1 + π¦1 β€ π1 and π2 β€ π΄2 β π¦2 <π΄2+π¦2 β€ π2, so thatπ is defined on [π΄1βπ¦1, π΄1+π¦1] and [π΄2βπ¦2, π΄2 + π¦2]. The properties of this mapping are embodied inthe following theorem.
Theorem 7. Suppose that π1, π2, π1, and π2 are positive realnumbers and
π΄ π = ππππ + ππππππ + ππ ,π¦π > 0,
where π¦π β€ ππ β ππππ + ππ min (ππ, ππ) for π = 1, 2.(25)
Also let π : Ξ2 β R be convex on the coordinates on Ξ2.(i) Themappingπ» is convex on the coordinates on [0, 1]2.(ii) We have the bounds
sup(π’,V)β[0,1]2
π»(π’, V)
= 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2= π» (1, 1) ,inf(π’,V)β[0,1]2
π»(π’, V) = π (π΄1, π΄2) = π» (0, 0) .
(26)
(iii) The mapping π» is monotonic nondecreasing on thecoordinates.
Proof. (i) Fix V β [0, 1], for all π’1, π’2 β [0, 1] and πΌ, π½ β₯ 0with πΌ + π½ = 1; we have
π»(πΌπ’1 + π½π’2, V) = 14π¦1π¦2
β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (πΌ (π’1π 1 + (1 β π’1) π΄1)β +π½ (π’2π 1 + (1 β π’2) π΄1) , Vπ 2+ (1 β V) π΄2) ππ 1 ππ 2 β€ πΌ
4π¦1π¦2β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π’1π 1 + (1 β π’1) π΄1, Vπ 2+ (1 β V) π΄2) ππ 1 ππ 2 + π½
4π¦1π¦2β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π’2π 1 + (1 β π’2) π΄1, Vπ 2+ (1 β V) π΄2) ππ 1 ππ 2 = πΌπ» (π’1, V)+ π½π» (π’2, V) .
(27)
If π’ β [0, 1] is fixed then, for all V1, V2 β [0, 1] and πΌ, π½ β₯ 0with πΌ + π½ = 1, we also haveπ»(π’, πΌV1 + π½V2) β€ πΌπ»(π’, V1) +π½π»(π’, V2) and the statement is proved.
(ii) Since π is convex on coordinates, we have by Jensen'sinequality for integrals that
π»(π’, V) β₯ 12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
π(π’π 1 + (1 β π’)π΄1, 12π¦2β β«π΄2+π¦2π΄2βπ¦2
(Vπ 2 + (1 β V) π΄2) ππ 2)ππ 1 = 12π¦1
β β«π΄1+π¦1π΄1βπ¦1
π (π’π 1 + (1 β π’)π΄1, π΄2) ππ 1
β₯ π( 12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
(π’π + (1 β π’)π΄1) ππ 1, π΄2)= π (π΄1, π΄2) .
(28)
By the convexity ofπ» on the coordinates, we have
π»(π’, V) β€ 12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
[V 12π¦2
β β«π΄2+π¦2π΄2βπ¦2
π (π’π 1 + (1 β π’)π΄1, π 2) ππ 2 + (1 β V) 12π¦2
β β«π΄2+π¦2π΄2βπ¦2
π (π’π 1 + (1 β π’)π΄1, π΄2) ππ 2] ππ 1
β€ V12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
[π’ 12π¦1
β β«π΄1+π¦1π΄1βπ¦1
π (π 1, π 2) ππ 1 + (1 β π’) 12π¦1
Journal of Mathematics 5
β β«π΄1+π¦1π΄1βπ¦1
π (π΄1, π 2) ππ 1] ππ 2 + (1 β V) 12π¦2
β β«π΄2+π¦2π΄2βπ¦2
[π’ 12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
π (π 1, π΄2) ππ 1 + (1 β π’)
β 12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
π (π΄1, π΄2) ππ 1] ππ 2 = π’V 14π¦1π¦2
β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2 + (1 β π’) V
β 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π΄1, π 2) ππ 1 ππ 2+ π’ (1 β V) 1
4π¦1π¦2β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π΄2) ππ 1 ππ 2 + (1 β π’)
β (1 β V) 14π¦1π¦2
β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π΄1, π΄2) ππ 1 ππ 2 = π’V 14π¦1π¦2
β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2 + (1 β π’) V 12π¦2
β β«π΄2+π¦2π΄2βπ¦2
π (π΄1, π 2) ππ 2 + π’ (1 β V) 12π¦1
β β«π΄1+π¦1π΄1βπ¦1
π (π 1, π΄2) ππ 1 + (1 β π’) (1 β V) π (π΄1, π΄2) .(29)
By Hadamardβs inequality introduced by Lupas (11), wealso have
π (π΄1, π 2) β€ 12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
π (π 1, π 2) ππ 1,
π (π 1, π΄2) β€ 12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
π (π 1, π 2) ππ 2.(30)
Thus by integration, we get that
12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
π (π΄1, π 2) ππ 2
β€ 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2,12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
π (π 1, π΄2) ππ 1
β€ 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2.
(31)
Also using the result
π (π΄1, π΄2) β€ 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2 (32)
we deduce the inequality
π»(π’, V)β€ [π’V + (1 β π’) V + π’ (1 β V) + (1 β π’) (1 β V)]β 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2 = 14π¦1π¦2
β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2,
(33)
for all (π’, V) β [0, 1]2 and the second bound in (ii) is proved.(iii) Firstly, we show thatπ»(π’, V) β₯ π»(0, V) for all [π’, V] β
[0, 1]2. By Jensenβs inequality for integrals, we haveπ»(π’, V) β₯ 1
2π¦2 β«π΄2+π¦2
π΄2βπ¦2
π( 12π¦1
β β«π΄1+π¦1π΄1βπ¦1
(π’π 1 + (1 β π’)π΄1) ππ 1, Vπ 2 + (1 β V)
β π΄2)ππ 2 = 12π¦2 β«
π΄2+π¦2
π΄2βπ¦2
π (π΄1, Vπ 2+ (1 β V) π΄2) ππ 2 = π» (0, V)
(34)
for all [π’, V] β [0, 1]2.Now let 0 β€ π’1 β€ π’2 β€ 1. By the convexity of mapping
π»(β , V) for all [π’, V] β [0, 1]2, we haveπ»(π’2, V) β π» (π’1, V)π’2 β π’1 β₯ π» (π’1, V) β π» (0, V)
π’1 β₯ 0. (35)
The following theorem also holds.
Theorem 8. Suppose that π1, π2, π1, and π2 are positive realnumbers and
π΄ π = ππππ + ππππππ + ππ ,π¦π > 0,
where π¦π β€ ππ β ππππ + ππ min (ππ, ππ) for π = 1, 2.(36)
Also let π : Ξ2 β R be convex on the coordinates on Ξ2.(i) The mappingπ» is convex on [0, 1]2.(ii) Define the mapping β : [0, 1] β R, β(π’) fl π»(π’, π’).
Then β is convex, monotonic nondecreasing on [0, 1]and one has the bounds
6 Journal of Mathematics
supπ’β[0,1]
β (π’) = β (1)
= 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2,infπ’β[0,1]
β (π’) = β (0) = π (π΄1, π΄2) .(37)
Proof. (i) Let (π’1, π’2), (V1, V2) β [0, 1]2 and πΌ, π½ β₯ 0 with πΌ +π½ = 1. Since π : Ξ2 β R is convex on Ξ2, we haveπ»(πΌ (π’1, π’2) + π½ (V1, V2)) = π» (πΌπ’1 + π½V1, πΌπ’2 + π½V2) = 1
4π¦1π¦2β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (πΌ (π’1π 1 + (1 β π’1) π΄1, π’2π 2 + (1 β π’2) π΄2)β +π½ (V1π 1 + (1 β V1) π΄1, V2π 2 + (1 β V2) π΄2)) ππ 1 ππ 2β€ πΌ β 1
4π¦1π¦2 β«π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π’1π 1 + (1 β π’1) π΄1, π’2π 2+ (1 β π’2) π΄2) ππ 1 ππ 2 + π½ β 1
4π¦1π¦2β β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (V1π 1 + (1 β V1) π΄1, V2π 2 + (1 β V2)β π΄2) ππ 1 ππ 2 = πΌπ» (π’1, π’2) + π½π» (V1, V2) ,
(38)
which showsπ» is convex on [0, 1]2.(ii) Let π’1, π’2 β [0, 1] and πΌ, π½ β₯ 0 with πΌ + π½ = 1.
β (πΌπ’1 + π½π’2) = π» (πΌπ’1 + π½π’2, πΌπ’1 + π½π’2)= π» (πΌ (π’1, π’1) + π½ (π’2, π’2))
β€ πΌπ» (π’1, π’1) + π½π» (π’2, π’2)= πΌβ (π’1) + π½β (π’2) ,
(39)
which shows the convexity of β on [0, 1].We have, by the above theorem, that
β (π’) = π» (π’, π’) β₯ π» (0, 0) = π (π΄1, π΄2) ,π’ β [0, 1] ,
β (π’) = π» (π’, π’) β€ π» (1, 1)= 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 1 ππ 2,π’ β [0, 1] ,
(40)
which proves the required bounds.
Now let 0 β€ π’1 β€ π’2 β€ 1. By the convexity of mapping β,we have that
β (π’2) β β (π’1)π’2 β π’1 β₯ β (π’1) β β (0)π’1 β₯ 0, (41)
and the theorem is proved.
Next, for positive real numbers π1, π2, π1, π2 and π΄ π =(ππππ + ππππ)/(ππ + ππ) and π¦π > 0 where π¦π β€ ((ππ β ππ)/(ππ +ππ))min(ππ, ππ) for π = 1, 2, then for mapping π : Ξ2 =[π1, π1] Γ [π2, π2] β R we shall consider the mapping οΏ½οΏ½ :[0, 1]2 β [0,β), which is given as
οΏ½οΏ½ (π’, V) fl 116π¦21π¦22 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π (π’π 1 + (1 β π’) π‘1, Vπ 2 + (1 β V) π‘2) ππ‘2 ππ 2 ππ‘1 ππ 1, (42)
where π1 β€ π΄1 β π¦1 < π΄1 + π¦1 β€ π1 and π2 β€ π΄2 β π¦2 <π΄2 + π¦2 β€ π2, so that π is defined on [π΄1 β π¦1, π΄1 + π¦1] and[π΄2 β π¦2, π΄2 + π¦2].The next theorem contains the main properties of this
mapping.
Theorem 9. Suppose that π1, π2, π1, and π2 are positive realnumbers and
π΄ π = ππππ + ππππππ + ππ ,π¦π > 0,
where π¦π β€ ππ β ππππ + ππ min (ππ, ππ) for π = 1, 2.(43)
Also let π : Ξ2 β R be convex on the coordinates on Ξ2.
(i) We have the equalities
οΏ½οΏ½ (π’ + 12 , V) = οΏ½οΏ½ (12 β π’, V) ,βπ’ β [0, 12] , V β [0, 1] ,
οΏ½οΏ½ (π’, V + 12) = οΏ½οΏ½ (π’, 12 β V) ,βπ’ β [0, 12] , V β [0, 1] ,
οΏ½οΏ½ (1 β π’, V) = οΏ½οΏ½ (π’, V) ,οΏ½οΏ½ (π’, 1 β V) = οΏ½οΏ½ (π’, V) ,
β (π’, V) β Ξ2.
(44)
Journal of Mathematics 7
(ii) οΏ½οΏ½ is convex on the coordinates.(iii) We have the bounds
inf(π’,V)β[0,1]2
οΏ½οΏ½ (π’, V) = οΏ½οΏ½ (12 ,12) =
116π¦21π¦22 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π(π 1 + π‘12 , π 2 + π‘22 ) ππ‘2 ππ 2 ππ‘1 ππ 1,
sup(π’,V)β[0,1]2
οΏ½οΏ½ (π’, V) = οΏ½οΏ½ (0, 0) = οΏ½οΏ½ (1, 1) = 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 2 ππ 1.(45)
(iv) The mapping οΏ½οΏ½(β , V) is monotonic nonincreasing on[0, 1/2) and nondecreasing on [1/2, 1] for all V β [0, 1].A similar property has the mapping οΏ½οΏ½(π’, β ) for all π’ β[0, 1].
(v) We have the inequality
οΏ½οΏ½ (π’, V) β₯ max {π» (π 1, π 2) ,π» (1 β π 1, π 2) ,π» (π 1, 1 β π 2) ,π» (1 β π 1, 1 β π 2)} .
(46)
Proof. (i) and (ii) are obvious.(iii) By the convexity of π in the first variable, we get that
12 [π (π’π 1 + (1 β π’) π‘1, Vπ 2 + (1 β V) π‘2)+ π ((1 β π’) π 1 + π’π‘1, Vπ 2 + (1 β V) π‘2)]β₯ π(π 1 + π‘12 , Vπ 2 + (1 β V) π‘2)
(47)
for all (π 1, π‘1) β [π1, π1]2, (π 2, π‘2) β [π2, π2]2, and (π’, V) β[0, 1]2. Integrating on [π΄1 β π¦1, π΄1 + π¦1]2, we get14π¦21 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄1+π¦1π΄1βπ¦1
π (π’π 1 + (1 β π’) π‘1, Vπ 2
+ (1 β V) π‘2) ππ 1 ππ‘1 β₯ 14π¦21
β β«π΄1+π¦1π΄1βπ¦1
β«π΄1+π¦1π΄1βπ¦1
π(π 1 + π‘12 , Vπ 2+ (1 β V) π‘2)ππ 1 ππ‘1.
(48)
Similarly,
14π¦22 β«
π΄2+π¦2
π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π(π 1 + π‘12 , Vπ 2 + (1 β V)
β π‘2)ππ 2 ππ‘2 β₯ 14π¦22
β β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π(π 1 + π‘12 , π 2 + π‘22 ) ππ 2 ππ‘2.
(49)
Now integrating this inequality on [π΄1 β π¦1, π΄1 + π¦1]2 andtaking into account the above inequality, we deduce
οΏ½οΏ½ (π’, V) β₯ 116π¦21π¦22 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π(π 1 + π‘12 , π 2 + π‘22 ) ππ‘2 ππ 2 ππ‘1 ππ 1 (50)
for (π’, V) β [0, 1]2. The first bound in (iii) is proved and thesecond bound goes on likewise.(iv)Themonotonicity of οΏ½οΏ½(β , π ) follows by a similar argumentfrom (iii).
(v) By Jensenβs integral inequality, we have successively for all(π’, V) β [0, 1]2 that
οΏ½οΏ½ (π’, V) β₯ 18π¦1π¦22 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π( 12π¦1 β«
π΄1+π¦1
π΄1βπ¦1
(π’π 1 + (1 β π’) π‘1) ππ‘1, Vπ 2 + (1 β V) π‘2)ππ‘2 ππ 2 ππ 1
= 18π¦1π¦22 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π (π’π 1 + (1 β π’)π΄1, Vπ 2 + (1 β V) π‘2) ππ‘2 ππ 2 ππ 1
β₯ 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π’π 1 + (1 β π’)π΄1, Vπ 2 + (1 β V) π΄2) ππ 2 ππ 1 = π» (π 1, π 2) .
(51)
8 Journal of Mathematics
Similarly, we can easily show οΏ½οΏ½(1 β π’, V) β₯ π»(1 β π 1, π 2),οΏ½οΏ½(π’, 1βV) β₯ π»(π 1, 1βπ 2), and οΏ½οΏ½(1βπ’, 1βV) β₯ π»(1βπ 1, 1βπ 2).In addition, as οΏ½οΏ½(π’, V) = οΏ½οΏ½(1 β π’, V) = οΏ½οΏ½(π’, 1 β V) =οΏ½οΏ½(1 β π’, 1 β V) for all (π’, V) β [0, 1]2, so we deduce οΏ½οΏ½(π’, V) β₯
max{π»(π 1, π 2),π»(1 β π 1, π 2),π»(π 1, 1 β π 2),π»(1 β π 1, 1 β π 2)}.The theorem is thus proved.
Theorem 10. Suppose that π1, π2, π1, and π2 are positive realnumbers and
π΄ π = ππππ + ππππππ + ππ ,
π¦π > 0,where π¦π β€ ππ β ππππ + ππ min (ππ, ππ) for π = 1, 2.
(52)
Also let π : Ξ2 β R be convex on the coordinates on Ξ2.(i) The mapping οΏ½οΏ½ is convex on Ξ2.(ii) Define the mapping β : [0, 1] β R, β(π’) fl οΏ½οΏ½(π’, π’).
Then β is convex, monotonic nonincreasing on [0, 1/2]and nondecreasing on [1/2, 1] and one has the bounds
infπ’β[0,1]
β (π’) = β (12) =1
16π¦21π¦22 β«π΄1+π¦1
π΄1βπ¦1
β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π(π 1 + π‘12 , π 2 + π‘22 ) ππ‘2 ππ 2 ππ‘1 ππ 1,
supπ’β[0,1]
β (π’) = β (1) = β (0) = 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 2 ππ 1.(53)
(iii) One has the inequality
β (π’) β₯ max {β (π’) , β (1 β π’)} . (54)
Proof. (i) Let (π’1, π’2), (V1, V2) β [0, 1] and πΌ, π½ β₯ 0 with πΌ +π½ = 1. Since π is convex on Ξ2, we have
οΏ½οΏ½ (πΌ (π’1, π’2) + π½ (V1, V2)) = οΏ½οΏ½ (πΌπ’1 + π½V1, πΌπ’2 + π½V2) = 116π¦21π¦22
β β«π΄1+π¦1π΄1βπ¦1
β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π (πΌ (π’1π 1 + (1 β π’1) π‘1, π’2π 2 + (1 β π’2) π‘2) +π½ (V1π 1 + (1 β V1) π‘1, V2π 2 + (1 β V2) π‘2)) ππ‘2 ππ 2 ππ‘1 ππ 1
β€ πΌ β 116π¦21π¦22 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π (π’1π 1 + (1 β π’1) π‘1, π’2π 2 + (1 β π’2) π‘2) ππ‘2 ππ 2 ππ‘1 ππ 1 + π½ β 116π¦21π¦22
β β«π΄1+π¦1π΄1βπ¦1
β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π (V1π 1 + (1 β V1) π‘1, V2π 2 + (1 β V2) π‘2) ππ‘2 ππ 2 ππ‘1 ππ 1 = πΌοΏ½οΏ½ (π’1, π’2) + π½οΏ½οΏ½ (V1, V2) ,
(55)
which shows οΏ½οΏ½ is convex on Ξ2.(ii) Let π’1, π’2, β [0, 1] and πΌ, π½ β₯ 0 with πΌ + π½ = 1.
β (πΌπ’1 + π½π’2) = οΏ½οΏ½ (πΌπ’1 + π½π’2, πΌπ’1 + π½π’2)= οΏ½οΏ½ (πΌ (π’1, π’1) + π½ (π’2, π’2))
β€ πΌοΏ½οΏ½ (π’1, π’1) + π½οΏ½οΏ½ (π’2, π’2)= πΌβ (π’1) + π½β (π’2) ,
(56)
which shows the convexity of β on [0, 1].As
β (π’) = οΏ½οΏ½ (π’, π’) β₯ οΏ½οΏ½ (12 ,12) = β (
12) =
116π¦21π¦22 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄1+π¦1π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
β«π΄2+π¦2π΄2βπ¦2
π(π 1 + π‘12 , π 2 + π‘22 ) ππ‘2 ππ 2 ππ‘1 ππ 1,
β (π’) = οΏ½οΏ½ (π’, π’) β€ οΏ½οΏ½ (0, 0) = οΏ½οΏ½ (0, 0) = 14π¦1π¦2 β«
π΄1+π¦1
π΄1βπ¦1
β«π΄2+π¦2π΄2βπ¦2
π (π 1, π 2) ππ 2 ππ 1,(57)
which proves the required bounds.
Journal of Mathematics 9
It is obvious that β ismonotonic nonincreasing on [0, 1/2]and nondecreasing on [1/2, 1].
(iii) As we know
β (π’) = οΏ½οΏ½ (π’, π’) β₯ max {π» (π’, π’) ,π» (1 β π’, 1 β π’)}= max {β (π’) , β (1 β π’)} (58)
and the theorem is proved.
Competing Interests
The authors declare that there is no conflict of interestsregarding the publication of this paper.
References
[1] D. S. Mitrinovic and I. B. Lackovic, βHermite and convexity,βAequationes Mathematicae, vol. 28, no. 3, pp. 229β232, 1985.
[2] A. M. Fink, βA best possible Hadamard inequality,βMathemati-cal Inequalities and Applications, vol. 1, no. 2, pp. 223β230, 1998.
[3] S. S. Dragomir, βA mapping in connection to Hadamardβsinequalities,β Osterreichische Akademie der Wissenschaften.Mathematisch-Naturwissenschaftliche, vol. 128, pp. 17β20, 1991.
[4] S. S. Dragomir, βOn Hadamardβs inequalities for convex func-tions,βMathematica Balkanica. New Series, vol. 6, no. 3, pp. 215β222, 1992.
[5] S. S. Dragomir, D. S. Milosevic, and J. Sandor, βOn some refine-ments of Hadamardβs inequalities and applications,β PublikacijeElektrotehnickog Fakulteta. Serija Matematika, vol. 4, pp. 3β10,1993.
[6] G. Farid, M. Marwan, and A. U. Rehman, βFejer-hadamardinequality for convex functions on the coordinates in a rectanglefrom the plane,β International Journal of Analysis and Applica-tions, vol. 10, no. 1, pp. 40β47, 2016.
[7] G. Farid and A. Ur Rehman, βGeneralization of the Fejer-Hadamardβs Inequality for convex function on coordinates,βKoreanMathematical Society. Communications, vol. 31, no. 1, pp.53β64, 2016.
[8] G. Farid, M. Marwan, and A. U. Rehman, βNew mean valuetheorems and generalization of Hadamard inequality via coor-dinated m-convex functions,β Journal of Inequalities and Appli-cations, vol. 2015, article 283, pp. 1β11, 2015.
[9] S. S. Dragomir and C. E. M. Pearce, Selected Topics on Hermite-Hadamard Inequalities, (RGMIAMonographs), VictoriaUniver-sity Press, 2000.
[10] S. S. Dragomir, βOn the Hadamards inequality for convexfunction on the co-ordinates in a rectangle from the plane,βTaiwanese Journal of Mathematics, vol. 5, no. 4, pp. 775β778,2001.
[11] A. Lupas, βA generalization of Hadamard inequalities forconvex functions,β Publikacije Elektrotehnickog Fakulteta. SerijaMatematika I Fizika, vol. 544β576, pp. 115β121, 1976.
[12] J. E. Pecaric, F. Proschan, and Y. Tong,Convex Functions, PartialOrderings, and Statistical Applications, vol. 187, Academic Press,New York, NY, USA, 1992.
Submit your manuscripts athttp://www.hindawi.com
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical Problems in Engineering
Hindawi Publishing Corporationhttp://www.hindawi.com
Differential EquationsInternational Journal of
Volume 2014
Applied MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Probability and StatisticsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Mathematical PhysicsAdvances in
Complex AnalysisJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
OptimizationJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
CombinatoricsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Operations ResearchAdvances in
Journal of
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Function Spaces
Abstract and Applied AnalysisHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
International Journal of Mathematics and Mathematical Sciences
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Algebra
Discrete Dynamics in Nature and Society
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Decision SciencesAdvances in
Discrete MathematicsJournal of
Hindawi Publishing Corporationhttp://www.hindawi.com
Volume 2014 Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014
Stochastic AnalysisInternational Journal of