It is known that an efficient method for interpolation of very large scattered data sets is the method of Shepard. Unfortunately, it reproduces only the constants. In this paper, we first generalize an expansion in bivariate even order Bernoulli polynomials for real functions possessing a sufficient number of derivatives. Finally, by combining the known Shepard operator with the even order Bernoulli bivariate operator, we construct a kind of new approximated operator satisfying the higher order polynomial reproducibility. We study this combined operator and give some error bounds in terms of the modulus of continuity of high order and also with Peano's theorem. Numerical comparisons show that this new technique provides the higher degree of accuracy. Furthermore, the advantage of our method is that the algorithm is very simple and easy to implement.
Citation: Ruifeng Wu. A kind of even order Bernoulli-type operator with bivariate Shepard[J]. AIMS Mathematics, 2023, 8(7): 15299-15316. doi: 10.3934/math.2023782
[1] | Ahmed A. El-Deeb, Osama Moaaz, Dumitru Baleanu, Sameh S. Askar . A variety of dynamic α-conformable Steffensen-type inequality on a time scale measure space. AIMS Mathematics, 2022, 7(6): 11382-11398. doi: 10.3934/math.2022635 |
[2] | Mohammed S. El-Khatib, Atta A. K. Abu Hany, Mohammed M. Matar, Manar A. Alqudah, Thabet Abdeljawad . On Cerone's and Bellman's generalization of Steffensen's integral inequality via conformable sense. AIMS Mathematics, 2023, 8(1): 2062-2082. doi: 10.3934/math.2023106 |
[3] | Ahmed A. El-Deeb, Dumitru Baleanu, Nehad Ali Shah, Ahmed Abdeldaim . On some dynamic inequalities of Hilbert's-type on time scales. AIMS Mathematics, 2023, 8(2): 3378-3402. doi: 10.3934/math.2023174 |
[4] | Ahmed A. El-Deeb, Samer D. Makharesh, Sameh S. Askar, Dumitru Baleanu . Bennett-Leindler nabla type inequalities via conformable fractional derivatives on time scales. AIMS Mathematics, 2022, 7(8): 14099-14116. doi: 10.3934/math.2022777 |
[5] | Elkhateeb S. Aly, Y. A. Madani, F. Gassem, A. I. Saied, H. M. Rezk, Wael W. Mohammed . Some dynamic Hardy-type inequalities with negative parameters on time scales nabla calculus. AIMS Mathematics, 2024, 9(2): 5147-5170. doi: 10.3934/math.2024250 |
[6] | Marwa M. Ahmed, Wael S. Hassanein, Marwa Sh. Elsayed, Dumitru Baleanu, Ahmed A. El-Deeb . On Hardy-Hilbert-type inequalities with α-fractional derivatives. AIMS Mathematics, 2023, 8(9): 22097-22111. doi: 10.3934/math.20231126 |
[7] | Bingxian Wang, Mei Xu . Asymptotic behavior of some differential inequalities with mixed delays on time scales and their applications. AIMS Mathematics, 2024, 9(6): 16453-16467. doi: 10.3934/math.2024797 |
[8] | Jian-Mei Shen, Saima Rashid, Muhammad Aslam Noor, Rehana Ashraf, Yu-Ming Chu . Certain novel estimates within fractional calculus theory on time scales. AIMS Mathematics, 2020, 5(6): 6073-6086. doi: 10.3934/math.2020390 |
[9] | Awais Younus, Khizra Bukhsh, Manar A. Alqudah, Thabet Abdeljawad . Generalized exponential function and initial value problem for conformable dynamic equations. AIMS Mathematics, 2022, 7(7): 12050-12076. doi: 10.3934/math.2022670 |
[10] | Haytham M. Rezk, Mohammed Zakarya, Amirah Ayidh I Al-Thaqfan, Maha Ali, Belal A. Glalah . Unveiling new reverse Hilbert-type dynamic inequalities within the framework of Delta calculus on time scales. AIMS Mathematics, 2025, 10(2): 2254-2276. doi: 10.3934/math.2025104 |
It is known that an efficient method for interpolation of very large scattered data sets is the method of Shepard. Unfortunately, it reproduces only the constants. In this paper, we first generalize an expansion in bivariate even order Bernoulli polynomials for real functions possessing a sufficient number of derivatives. Finally, by combining the known Shepard operator with the even order Bernoulli bivariate operator, we construct a kind of new approximated operator satisfying the higher order polynomial reproducibility. We study this combined operator and give some error bounds in terms of the modulus of continuity of high order and also with Peano's theorem. Numerical comparisons show that this new technique provides the higher degree of accuracy. Furthermore, the advantage of our method is that the algorithm is very simple and easy to implement.
The renowned integral Steffensen's inequality [1] is written as: Let f and g be integrable functions on [a,b] such that f is nonincreasing and 0≤g(J)≤1 on [a,b]. Then
∫bb−λf(J)dJ≤∫baf(J)g(J)dJ≤∫a+λaf(J)dJ, | (1.1) |
where λ=∫bag(J)dJ.
It is simple to notice that inequalities (1.1) are reversed if f is nondecreasing.
The discrete version of the Steffensen inequality [2] states:
Theorem A. Assume that {f(k)}nk=1 is a nonincreasing nonnegative real sequence and {g(k)}nk=1 is a real sequence such that 0≤g(k)≤1 for every k. Furthermore, let that λ1, λ2∈{1,…,n} be such that λ2≤∑nk=1g(k)≤λ1. Then
n∑k=n−λ2+1f(k)≤n∑k=1f(k)g(k)≤λ1∑k=1f(k). | (1.2) |
Jakšetić et al. [3] established the following interesting results among many other similar results for a positive finite measure μ. States:
Theorem B. Let ˆμ be a positive finite measure on B([a,b]), and let f, g:[a,b→R be measurable functions on [a,b] such that f is nonincreasing and 0≤g(J)≤1 for all t∈[a,b]. Further, let ˆμ([c,d])=∫[a,b]g(J)dˆμ(J), where [c,d]⊆[a,b]. Then
∫[a,b]f(J)g(J)dˆμ(J)≤∫[c,d]f(J)g(J)dˆμ(J)+∫[a,c](f(J)−f(d))g(J)dˆμ(J). |
Also, the authors proved that:
Theorem C. Let f, g:[a,b]→R be measurable functions on [a,b] such that f is nonincreasing and 0≤g(J)≤1 for all t∈[a,b]. Further, let ˆμ([c,d])=∫[a,b]g(J)dˆμ(J), where [c,d]⊆[a,b]. If ˆμ is a positive finite measure on B([a,b]), then
∫[c,d]f(J)dˆμ(J)−∫[d,b](f(c)−f(J))g(J)dˆμ(J)≤∫[a,b]f(J)g(J)dˆμ(J). |
In 1982, Pečarić [4] gave speculation of the Steffensen inequality as the following two hypotheses.
Theorem 1.1. Let ˆf, ˆg, ˆh:[a,b]→R be integrable functions on [a,b] such that ˆf/ˆh is nonincreasing and ˆh is nonnegative. Further, let 0≤ˆg(J)≤1 ∀J∈[a,b]. Then
∫baˆf(J)ˆg(J)dJ≤∫a+ˆ℘aˆfdJ, | (1.3) |
where ˆ℘ is the solution of the equation
∫a+ˆ℘aˆh(J)dJ=∫baˆh(J)ˆg(J)dJ. |
We get the reverse of (1.3), if ^f(J)/^h(J) is nondecreasing.
Theorem 1.2. Let ˆf, ˆg, ˆh:[a,b]→R be integrable functions on [a,b] such that ˆf/ˆh is nonincreasing and ˆh is nonnegative. Further, let 0≤ˆg(J)≤1 ∀J∈[a,b]. Then
∫bb−ˆ℘ˆf(J)dJ≤∫baˆf(J)ˆg(J)dJ, | (1.4) |
where ˆ℘ gives us the solution of
∫bb−ˆ℘ˆh(J)dJ=∫baˆh(J)ˆg(J)dJ. | (1.5) |
We get the reverse of (1.4), if ^f(J)/^h(J) is nondecreasing.
Wu and Srivastava in [5] acquired the accompanying result.
Theorem 1.3. Let ˆf, ˆg, ˆh:[a,b]→R be integrable functions on [a,b] such that ˆf is nonincreasing. Further, let 0≤ˆg(J)≤ˆh(J) ∀J∈[a,b]. Then the following integral inequalities hold true:
∫bb−ˆ℘ˆf(J)ˆh(J)dJ≤∫bb−ˆ℘(ˆf(J)ˆh(J)−[ˆf(J)−ˆf(b−ˆ℘)][ˆh(J)−ˆg(J)])dJ≤∫baˆf(J)ˆg(J)dJ≤∫a+ˆ℘a(ˆf(J)ˆh(J)−[ˆf(J)−ˆf(a+ˆ℘)][ˆh(J)−ˆg(J)])dJ≤∫a+ˆ℘aˆf(J)ˆh(J)dJ, |
where ˆ℘ gives us the solution of
∫a+ˆ℘aˆh(J)dJ=∫baˆg(J)dJ=∫bb−ˆ℘ˆh(J)dJ. |
The following interesting findings were published in [6].
Theorem 1.4. Suppose the integrability of ˆg, ˆh, ˆf, ψ:[a,b]→R such that ˆf is nonincreasing. Also suppose 0≤ˆψ(J)≤ˆg(J)≤ˆh(J)−ˆψ(J) for all J∈[a,b]. Then
∫baˆf(J)ˆg(J)dJ≤∫a+ˆ℘aˆf(J)ˆh(J)dJ−∫ba|(ˆf(J)−ˆf(a+ˆ℘))ψ(J)|dJ, |
where ˆ℘ is given by
∫a+ˆ℘aˆh(J)dJ=∫baˆg(J)dJ. |
Theorem 1.5. Under the hypotheses of Theorem 1.4. Then
∫bb−ˆ℘ˆf(J)ˆh(J)dJ+∫ba|(ˆf(J)−ˆf(b−ˆ℘))ˆψ(J)|dJ≤∫baˆf(J)ˆg(J)dJ, |
where ˆ℘ is given by
∫bb−ˆ℘ˆh(J)dJ=∫baˆg(J)dJ. |
The calculus of time scales with the intention to unify discrete and continuous analysis (see [7]) was proposed by Hilger [8]. For more details on the time scales calculus we refer to the book by Bohner and Peterson [9].
Lately, several dynamic inequalities on time scales has been investigated by using exclusive authors who have been inspired with the aid of a few applications (see[10,11,12,13,14,15,16,17,18]). Some authors created different results regarding fractional calculus on time scales to provide associated dynamic inequalities (see [19,20,21,22,23,24,25,26,27]).
In this article, we explore new generalizations of the integral Steffensen inequality given in [4,5,6] via general time scale measure space with a positive σ-finite measure. We also retrieve some of the integral inequalities known in the literature as special cases of our tests.
In what follows B([a,b]T) is Borel σ-algebra [a,b]. Next, we enroll the accompanying suppositions for the verifications of our primary outcomes:
(S1) ([a,b]T,B([a,b]T),ˆμ) is time scale measure space with a positive σ-finite measure on B([a,b]T).
(S2) η, Υ, Ξ:[a,b]T→R are Δˆμ-integrable functions on [a,b]T.
(S3) η/Ξ is nonincreasing and Ξ is nonnegative.
(S4) 0≤Υ(J)≤1 for all J∈[a,b]T.
(S5) ˆ℘∈[0,∞).
(S6) η is nonincreasing.
(S7) 1≤Υ(J)≤Ξ(J) for all J∈[a,b]T.
(S8) 0≤ψ(J)≤Υ(J)≤Ξ(J)−ψ(J) for all J∈[a,b]T.
(S9) 0≤M≤Υ(J)≤1−M for all J∈[a,b]T.
(S10) 0≤ψ(J)≤Υ(J)≤1−ψ(J) for all J∈[a,b]T.
ˆ℘ is the solution of the equations listed below:
(S11) ∫[a,a+ˆ℘]TΞ(J)Δˆμ=∫[a,b]TΞ(J)Υ(J)Δˆμ.
(S12) ∫[b−ˆ℘,b]TΞ(J)Δˆμ=∫[a,b]TΞ(J)Υ(J)Δˆμ.
(S13) ∫[a,a+ˆ℘]TΞ(J)Δˆμ=∫[a,b]TΥ(J)Δˆμ=∫[b−ˆ℘,b]TΞ(J)Δˆμ.
(S14) ∫[a,a+ˆ℘]TΞ(J)Δˆμ=∫[a,b]TΥ(J)Δˆμ.
(S15) ∫[b−ˆ℘,b]TΞ(J)Δˆμ=∫[a,b]TΥ(J)Δˆμ.
Presently, we are prepared to state and explain the principle results that make bigger numerous effects inside the literature.
Theorem 2.1. Let S1, S2, S3, S4, S5 and S11 be satisfied. Then
∫[a,b]Tη(J)Υ(J)Δˆμ≤∫[a,a+ˆ℘]Tη(J)Δˆμ. | (2.1) |
We get the reverse of (2.1), if η/Ξ is nondecreasing.
Proof. From our hypotheses, we observe that,
∫[a,a+ˆ℘]Tη(J)Δˆμ−∫[a,b]Tη(J)Υ(J)Δˆμ=∫[a,a+ˆ℘]TΞ(J)[1−Υ(J)]η(J)Ξ(J)Δˆμ−∫[a+ˆ℘,b]Tη(J)Υ(J)Δˆμ≥η(a+ˆ℘)Ξ(a+ˆ℘)∫[a,a+ˆ℘]TΞ(J)[1−Υ(J)]Δˆμ−∫[a+ˆ℘,b]Tη(J)Υ(J)Δˆμ=η(a+ˆ℘)Ξ(a+ˆ℘)[∫[a,a+ˆ℘]TΞ(J)Δˆμ−∫[a,a+ˆ℘]TΞ(J)Υ(J)Δˆμ]−∫[a+ˆ℘,b]Tη(J)Υ(J)Δˆμ=η(a+ˆ℘)Ξ(a+ˆ℘)[∫[a,b]TΞ(J)Υ(J)Δˆμ−∫[a,a+ˆ℘]TΞ(J)Υ(J)Δˆμ]−∫[a+ˆ℘,b]Tη(J)Υ(J)Δˆμ=η(a+ˆ℘)Ξ(a+ˆ℘)∫[a+ˆ℘,b]TΞ(J)Υ(J)Δˆμ−∫[a+ˆ℘,b]Tη(J)Υ(J)Δˆμ=∫[a+ˆ℘,b]TΞ(J)Υ(J)(η(a+ˆ℘)Ξ(a+ˆ℘)−η(J)Ξ(J))Δˆμ≥0. |
The proof is complete.
Remark 2.1. In case of T=R and related to Lebesgue measure in Theorem 2.1, we recollect [4,Theorem 1].
Theorem 2.2. Assumptions S1, S2, S3, S4, S5 and S12 imply
∫[b−ˆ℘,b]Tη(J)Δˆμ≤∫[a,b]Tη(J)Υ(J)Δˆμ. | (2.2) |
We get the reverse of (2.2), if η/Ξ is nondecreasing.
Proof. From our hypotheses, we observe that,
∫[b−ˆ℘,b]Tη(J)Δˆμ−∫[a,b]Tη(J)Υ(J)Δˆμ=∫[b−ˆ℘,b]TΞ(J)[1−Υ(J)]η(J)Ξ(J)Δˆμ−∫[a,b−ˆ℘]Tη(J)Υ(J)Δˆμ≤η(b−ˆ℘)Ξ(b−ˆ℘)∫[b−ˆ℘,b]TΞ(J)[1−Υ(J)]Δˆμ−∫[a,b−ˆ℘]Tη(J)Υ(J)Δˆμ=η(b−ˆ℘)Ξ(b−ˆ℘)[∫[b−ˆ℘,b]TΞ(J)Δˆμ−∫[b−ˆ℘,b]TΞ(J)Υ(J)Δˆμ]−∫[a,b−ˆ℘]Tη(J)Υ(J)Δˆμ=η(b−ˆ℘)Ξ(b−ˆ℘)[∫[a,b]TΞ(J)Υ(J)Δˆμ−∫[b−ˆ℘,b]TΞ(J)Υ(J)Δˆμ]−∫[a,b−ˆ℘]Tη(J)Υ(J)Δˆμ=η(b−ˆ℘)Ξ(b−ˆ℘)∫[a,b−ˆ℘]TΞ(J)Υ(J)Δˆμ−∫[a,b−ˆ℘]Tη(J)Υ(J)Δˆμ=∫[a,b−ˆ℘]TΞ(J)Υ(J)(η(b−ˆ℘)Ξ(b−ˆ℘)−η(J)Ξ(J))Δˆμ≤0. |
Remark 2.2. By observing Lebesgue measure in Theorem 2.2, and T=R, we recapture [4,Theorem 2].
We will need the following lemma to prove the subsequent results.
Lemma 2.1. Let S1, S2, S5 hold, such that
∫[a,a+ˆ℘]TΞ(J)Δˆμ=∫[a,b]TΥ(J)Δˆμ=∫[b−ˆ℘,b]TΞ(J)Δˆμ. |
Then
∫[a,b]Tη(J)Υ(J)Δˆμ=∫[a,a+ˆ℘]T(η(J)Ξ(J)−[η(J)−η(a+ˆ℘)][Ξ(J)−Υ(J)])Δˆμ+∫[a+ˆ℘,b]T[η(J)−η(a+ˆ℘)]Υ(J)Δˆμ, | (2.3) |
and
∫[a,b]Tη(J)Υ(J)Δˆμ=∫[a,b−ˆ℘]T[η(J)−η(b−ˆ℘)]Υ(J)Δˆμ+∫[b−ˆ℘,b]T(η(J)Ξ(J)−[η(J)−η(b−ˆ℘)][Ξ(J)−Υ(J)])Δˆμ. | (2.4) |
Proof. The suppositions of the Lemma imply that
a≤a+ˆ℘≤banda≤b−ˆ℘≤b. |
Firstly, we prove the validity of the integral identity (2.3). Indeed, by direct computation, and from our hypotheses, we find that
∫[a,a+ˆ℘]T(η(J)Ξ(J)−[η(J)−η(a+ˆ℘)][Ξ(J)−Υ(J)])Δˆμ−∫[a,b]Tη(J)Υ(J)Δˆμ=∫[a,a+ˆ℘]T(η(J)Ξ(J)−η(J)Υ(J)−[η(J)−η(a+ˆ℘)][Ξ(J)−Υ(J)])Δˆμ+∫[a,a+ˆ℘]Tη(J)Υ(J)Δˆμ−∫[a,b]Tη(J)Υ(J)Δˆμ=∫[a,a+ˆ℘]Tη(a+ˆ℘)[Ξ(J)−Υ(J)]Δˆμ−∫[a+ˆ℘,b]Tη(J)Υ(J)Δˆμ=η(a+ˆ℘)(∫[a,a+ˆ℘]TΞ(J)Δˆμ−∫[a,a+ˆ℘]TΥ(J)Δˆμ)−∫[a+ˆ℘,b]Tη(J)Υ(J)Δˆμ. | (2.5) |
Since
∫[a,a+ˆ℘]TΞ(J)Δˆμ=∫[a,b]TΥ(J)Δˆμ, |
we have
η(a+ˆ℘)(∫[a,a+ˆ℘]TΞ(J)Δˆμ−∫[a,a+ˆ℘]TΥ(J)Δˆμ)−∫[a+ˆ℘,b]Tη(J)Υ(J)Δˆμ=η(a+ˆ℘)(∫[a,b]TΥ(J)Δˆμ−∫[a,a+ˆ℘]TΥ(J)Δˆμ)−∫[a+ˆ℘,b]Tη(J)Υ(J)Δˆμ=η(a+ˆ℘)∫[a+ˆ℘,b]TΥ(J)Δˆμ−∫[a+ˆ℘,b]Tη(J)Υ(J)Δˆμ)=∫[a+ˆ℘,b]T[η(a+ˆ℘)−η(J)]Υ(J)Δˆμ. | (2.6) |
Combination of (2.5) and (2.6) led to the required integral identity (2.3) asserted by the Lemma. The integral identity (2.4) can be proved similarly. The proof is done.
Theorem 2.3. Suppose S1, S2, S5, S6, S7 and S13 give
∫[b−ˆ℘,b]Tη(J)Ξ(J)Δˆμ≤∫[b−ˆ℘,b]T(η(J)Ξ(J)−[η(J)−η(b−ˆ℘)][Ξ(J)−Υ(J)])Δˆμ≤∫[a,b]Tη(J)Υ(J)Δˆμ≤∫[a,a+ˆ℘]T(η(J)Ξ(J)−[η(J)−η(a+ˆ℘)][Ξ(J)−Υ(J)])Δˆμ≤∫[a,a+ˆ℘]Tη(J)Ξ(J)Δˆμ. |
Proof. From our hypotheses. In perspective of the considerations that the function η is nonincreasing on [a,b] and 0≤Υ(J)≤Ξ(J) for all J∈[a,b], we infer that
∫[a,b−ˆ℘]T[η(J)−η(b−ˆ℘)]Υ(J)Δˆμ≥0, | (2.7) |
and
∫[b−ˆ℘,b]T[η(b−ˆ℘)−η(J)][Ξ(J)−Υ(J)]Δˆμ≥0. | (2.8) |
Using (2.3), (2.7) and (2.8), we find that
∫[b−ˆ℘,b]Tη(J)Ξ(J)Δˆμ≤∫[b−ˆ℘,b]T(η(J)Ξ(J)−[η(J)−η(b−ˆ℘)][Ξ(J)−Υ(J)])Δˆμ≤∫[a,b]Tη(J)Υ(J)Δˆμ. | (2.9) |
In the same way as above, we can prove that
∫[a,b]Tη(J)Υ(J)Δˆμ≤∫[a,a+ˆ℘]T(η(J)Ξ(J)−[η(J)−η(a+ˆ℘)][Ξ(J)−Υ(J)])Δˆμ≤∫[a,a+ˆ℘]Tη(J)Ξ(J)Δˆμ, | (2.10) |
The confirmation is finished by joining the integral inequalities (2.9) and (2.10).
Remark 2.3. We can reclaim [5,Theorem 1] with the use of Lebesgue measure in Theorem 2.3, and T=R.
Theorem 2.4. Assume S1, S2, S5, S6, S8 and S13 be fulfilled. Then
∫[b−ˆ℘,b]Tη(J)Ξ(J)Δˆμ+∫[a,b]T|[η(J)−η(b−ˆ℘)]ψ(J)|Δˆμ≤∫[a,b]Tη(J)Υ(J)Δˆμ≤∫[a,a+ˆ℘]Tη(J)Ξ(J)Δˆμ−∫[a,b]T|[η(J)−η(a+ˆ℘)]ψ(J)|Δˆμ. | (2.11) |
Proof. From our hypotheses. Clearly function η is nonincreasing on [a,b] and 0≤ψ(J)≤Υ(J)≤Ξ(J)−ψ(J) for all J∈[a,b], we obtain
∫[a,a+ˆ℘]T[η(J)−η(a+ˆ℘)][Ξ(J)−Υ(J)]Δˆμ+∫[a+ˆ℘,b]T[η(a+ˆ℘)−η(J)]Υ(J)Δˆμ=∫[a,a+ˆ℘]T|η(J)−η(a+ˆ℘)|[Ξ(J)−Υ(J)]Δˆμ+∫[a+ˆ℘,b]T|η(a+ˆ℘)−η(J)|Υ(J)Δˆμ≥∫[a,a+ˆ℘]T|η(J)−η(a+ˆ℘)|ψ(J)Δˆμ+∫[a+ˆ℘,b]T|η(a+ˆ℘)−η(J)|ψ(J)Δˆμ≥∫[a,b]T|[η(J)−η(a+ˆ℘)]ψ(J)|Δˆμ. |
Also
∫[a,a+ˆ℘]T[η(J)−η(a+ˆ℘)][Ξ(J)−Υ(J)]Δˆμ+∫[a+ˆ℘,b]T[η(a+ˆ℘)−η(J)]Υ(J)Δˆμ≥∫[a,b]T|[η(J)−η(a+ˆ℘)]ψ(J)|Δˆμ. | (2.12) |
Similarly, we find that
∫[a,b−ˆ℘]T[η(J)−η(b−ˆ℘)]Υ(J)Δˆμ+∫[b−ˆ℘,b]T[η(b−ˆ℘)−η(J)][Ξ(J)−Υ(J)]Δˆμ≥∫[a,b]T|[η(J)−η(b−ˆ℘)]ψ(J)|Δˆμ. | (2.13) |
By combining (2.3), (2.4), (2.12) and (2.13), we arrive at the inequality (2.11) asserted by Theorem 2.
Remark 2.4. If we take T=R, and consider the Lebesgue measure in Theorem 2.4, we recapture [5,Theorem 2].
In the following theorem, we use the additional parameters ˆ℘1, ˆ℘2∈[0,∞).
Theorem 2.5. Let S1, S2, S5, S6, S9 be satisfied, and
0≤ˆ℘1≤∫[a,b]TΥ(J)Δˆμ≤ˆ℘2≤b−a. |
Then
∫[b−ˆ℘1,b]Tη(J)Δˆμ+η(b)(∫[a,b]TΥ(J)Δˆμ−ˆ℘1)+M∫[a,b]T|η(J)−f(b−∫[a,b]TΥ(J)Δˆμ)|Δˆμ≤∫[a,b]Tη(J)Υ(J)Δˆμ≤∫[a,a+ˆ℘2]Tη(J)Δˆμ−η(b)(ˆ℘2−∫[a,b]TΥ(J)Δˆμ)−M∫[a,b]T|η(J)−f(a+∫[a,b]TΥ(J)Δˆμ)|Δˆμ. | (2.14) |
Proof. By using straightforward calculations, we have
∫[a,b]Tη(J)Υ(J)Δˆμ−∫[a,a+ˆ℘2]Tη(J)Δˆμ+η(b)(ˆ℘2−∫[a,b]TΥ(J)Δˆμ)=∫[a,b]Tη(J)Υ(J)Δˆμ−∫[a,a+ˆ℘2]Tη(J)Δˆμ+∫[a,a+ˆ℘2]Tη(b)Δˆμ−∫[a,b]Tη(b)Υ(J)Δˆμ=∫[a,b]T[η(J)−η(b)]Υ(J)Δˆμ−∫[a,a+ˆ℘2]T[η(J)−η(b)]Δˆμ≤∫[a,b]T[η(J)−η(b)]Υ(J)Δˆμ−∫[a,a+∫[a,b]TΥ(J)Δˆμ][η(J)−η(b)]Δˆμ, | (2.15) |
where we used the theorem's hypotheses
a≤a+ˆ℘1≤a+∫[a,b]TΥ(J)Δˆμ≤a+ˆ℘2≤b, |
and
η(J)−η(b)≥0for allJ∈[a,b]. |
The function η(J)−η(b) is nonincreasing and integrable on [a,b] and by applying Theorem 2 with Ξ(J)=1, ψ(J)=M and η(J) replaced by η(J)−η(b), hence
∫[a,b]T[η(J)−η(b)]Υ(J)Δˆμ−∫[a,a+∫[a,b]TΥ(J)Δˆμ][η(J)−η(b)]Δˆμ≤−M∫[a,b]T|η(J)−f(a+∫[a,b]TΥ(J)Δˆμ)|Δˆμ. | (2.16) |
From (2.15) and (2.16) we obtain
∫[a,b]Tη(J)Υ(J)Δˆμ−∫[a,a+ˆ℘2]Tη(J)Δˆμ+η(b)(ˆ℘2−∫[a,b]TΥ(J)Δˆμ)≤−M∫[a,b]T|η(J)−f(a+∫[a,b]TΥ(J)Δˆμ)|Δˆμ, | (2.17) |
which is the right-hand side inequality in (2.14).
Similarly, one can show that
∫[a,b]Tη(J)Υ(J)Δˆμ−∫[b−ˆ℘1,b]Tη(J)Δˆμ+η(b)(∫[a,b]TΥ(J)Δˆμ−ˆ℘2)≥∫[a,b]T[η(J)−η(b)]Υ(J)Δˆμ+∫[b−∫[a,b]TΥ(J)Δˆμ,b][η(b)−η(J)]Δˆμ≥M∫[a,b]T|η(J)−f(b−∫[a,b]TΥ(J)Δˆμ)|Δˆμ, | (2.18) |
which is the left-hand side inequality in (2.14).
Remark 2.5. [5,Theorem 3] can be obtained if T=R and Lebesgue measure in Theorem 2.5.
Theorem 2.6. If S1, S2, S5, S6, S7 and S14 hold. Then
∫[a,b]Tη(J)Υ(J)Δˆμ≤∫[a,a+ˆ℘]Tη(J)Ξ(J)Δˆμ−∫[a,b]T|(η(J)−η(a+ˆ℘))ψ(J)|Δˆμ. | (2.19) |
Proof. Follows similar to the proof of the right-hand side inequality in Theorem 2.
Remark 2.6. If we take T=R, and consider the Lebesgue measure in Theorem 2.6, we recapture [6,Theorem 2.12].
Corollary 2.1. Hypotheses S1, S2, S3, S10 and S11 yield
∫[a,b]Tη(J)Υ(J)Δˆμ≤∫[a,a+ˆ℘]Tη(J)Δˆμ−∫[a,b]T|(η(J)Ξ(J)−η(a+ˆ℘)Ξ(a+ˆ℘))Ξ(J)ψ(J)|Δˆμ. | (2.20) |
Proof. Insert Υ(J)↦Ξ(J)Υ(J), η(J)↦η(J)/Ξ(J) and ψ(J)↦Ξ(J)ψ(J) in Theorem 2.
Remark 2.7. [6,Corollary 2.3] can be recovered with the help of T=R, and Lebesgue measure in Corollary 2.1.
Theorem 2.7. If S1, S2, S5, S6, S7 and S15 hold. Then
∫[b−ˆ℘,b]Tη(J)Ξ(J)Δˆμ+∫[a,b]T|(η(J)−η(b−ˆ℘))ψ(J)|Δˆμ≤∫[a,b]Tη(J)Υ(J)Δˆμ. | (2.21) |
Proof. Carry out the same proof of the left-hand side inequality in Theorem 2.
Remark 2.8. If we take T=R, and consider the Lebesgue measure in Theorem 2.7, we recapture [6,Theorem 2.13].
Corollary 2.2. Let S1, S2, S3, S9 and S12, be fulfilled. Then
∫[b−ˆ℘,b]Tη(J)Δˆμ+∫[a,b]T|(η(J)Ξ(J)−η(b−ˆ℘)Ξ(b−ˆ℘))Ξ(J)ψ(J)|Δˆμ≤∫[a,b]Tη(J)Υ(J)Δˆμ. | (2.22) |
Proof. Proof can be completed by taking Υ(J)↦Ξ(J)Υ(J), η(J)↦η(J)/Ξ(J) and ψ(J)↦Ξ(J)ψ(J) in Theorem 2.
Remark 2.9. By letting T=R, and consider the Lebesgue measure in Corollary 2.2, we recapture [6,Corollary 2.4].
In this article, we explore new generalizations of the integral Steffensen inequality given in [4,5,6] via general time scale measure space with a positive σ-finite measure, we generalize a number of those inequalities to a general time scale measure space. Besides that, in order to obtain some new inequalities as special cases, we also extend our inequalities to discrete and constant calculus.
This work was supported by Inho Hwang Incheon National University Research Grant 2021–2022.
The authors declare that there is no competing interest.
[1] | D. Shepard, A two-dimensional interpolation function for irregularly-spaced data, Proceedings of the 1968 23rd ACM National Conference, (1968), 517–524. https://doi.org/10.1145/800186.810616 |
[2] | Gh. Coman, L. Ţˆambulea, A Shepard-Taylor approximation formula, Studia Univ. Babeş-Bolyai Math., 33 (1998), 65–73. |
[3] | Gh. Coman, R. T. Trîmbiţaş, Combined Shepard univariate operators, East J. Approx., 7 (2001), 471–483. |
[4] |
R. Farwig, Rate of convergence of Shepard's global interpolation formula, Math. Comp., 46 (1986), 577–590. https://doi.org/10.1090/S0025-5718-1986-0829627-0 doi: 10.1090/S0025-5718-1986-0829627-0
![]() |
[5] | Gh. Coman, R.T. Trîmbiţaş, Shepard operators of Lagrange-type, Studia Univ. Babeş-Bolyai Math., 42 (1997), 75–83. |
[6] | Gh. Coman, Hermite-type Shepard operators, Rev. Anal. Numˊer. Thˊeor. Approx., 26 (1997), 33–38. |
[7] |
Gh. Coman, Shepard operators of Birkhoff-type, Calcolo, 35 (1998), 197–203. https://doi.org/10.1007/s100920050016 doi: 10.1007/s100920050016
![]() |
[8] |
F. Caira, F. Dell'Accio, Shepard-Bernoulli operators, Math. Comp., 76 (2007), 299–321. https://doi.org/10.1090/S0025-5718-06-01894-1 doi: 10.1090/S0025-5718-06-01894-1
![]() |
[9] |
T. C˘atinas, The bivariate Shepard operator of Bernoulli type, Calcolo, 44 (2007), 189–202. https://doi.org/10.1007/s10092-007-0136-x doi: 10.1007/s10092-007-0136-x
![]() |
[10] | T. Cˇatinaş, The combined Shepard-Lidstone bivariate operator, In: de Bruin, M.G. et al.(eds.) Trends an Applications in Constructive Approximation. International Series of Numerical Mathematics, 151 (2005), 77–89. https://doi.org/10.1007/3-7643-7356-3 |
[11] |
R. J. Renka, Multivariate Interpolation of Large Sets of Scattered Data, ACM Trans. Math. Software, 14 (1988), 139–148. https://doi.org/10.1145/45054.45055 doi: 10.1145/45054.45055
![]() |
[12] |
R. J. Renka, Algorithm 660, QSHEP2D: Quadratic Shepard Method for Bivariate Interpolation of Scattered Data, ACM Trans. Math. Software, 14 (1988), 149–150. https://doi.org/10.1145/45054.356231 doi: 10.1145/45054.356231
![]() |
[13] |
R. J. Renka, Algorithm 661, QSHEP3D: Quadratic Shepard Method for Trivariate Interpolation of Scattered Data, ACM Trans. Math. Software, 14 (1988), 151–152. https://doi.org/10.1145/45054.214374 doi: 10.1145/45054.214374
![]() |
[14] | M. G. Trîmbiţaş, Combined Shepard-least square operators-computing them using spatial data structures, Studia Univ. Babeş-Bolyai Math., 47 (2002), 119–128. |
[15] |
F. A. Costabile, F. Dell'Accio, F. Di Tommaso, Complementary Lidstone Interpolation on Scattered Data Sets, Numer. Algorithms, 67 (2013), 157–180. https://doi.org/10.1007/s11075-012-9659-6 doi: 10.1007/s11075-012-9659-6
![]() |
[16] |
R. Caira, F. Dell'Accio, F. Di Tommaso, On the bivariate Shepard-Lidstone operators, J. Comput. Appl. Math., 236 (2012), 1691–1707. https://doi.org/10.1016/j.cam.2011.10.001 doi: 10.1016/j.cam.2011.10.001
![]() |
[17] |
F. Dell'Accio, F. Di Tommaso, Complete Hermite-Birkhoff interpolation on scattered data by combined Shepard operators, J. Comput. Appl. Math., 300 (2016), 192–206. https://doi.org/10.1016/j.cam.2015.12.016 doi: 10.1016/j.cam.2015.12.016
![]() |
[18] |
F. Dell'Accio, F. Di Tommaso, Bivariate Shepard-Bernoulli operators, Math. Comput. Simulat., 141 (2017), 65–82. https://doi.org/10.1016/j.matcom.2017.07.002 doi: 10.1016/j.matcom.2017.07.002
![]() |
[19] |
O. Duman, B. Della Vecchia, Approximation to integrable functions by modified complex Shepard operators, J. Math. Anal. Appl., 512 (2022), 126161. https://doi.org/10.1016/j.jmaa.2022.126161 doi: 10.1016/j.jmaa.2022.126161
![]() |
[20] |
O. Duman, B. Della Vecchia, Complex Shepard operators and their summability, Results Math., 76 (2021), 214. https://doi.org/10.1007/s00025-021-01520-4 doi: 10.1007/s00025-021-01520-4
![]() |
[21] |
F. Dell'Accio, F. Di Tommaso, O. Nouisser, N. Siar, Solving Poisson equation with Dirichlet conditions through multinode Shepard operators, Comput. Math. Appl., 98 (2021), 254–260. https://doi.org/10.1016/j.camwa.2021.07.021 doi: 10.1016/j.camwa.2021.07.021
![]() |
[22] |
R. K. Beatson, M. J. D. Powell, Univariate multiquadric approximation: Quasi-interpolation to scattered data, Constr. Approx., 8 (1992), 275–288. https://doi.org/10.1007/BF01279020 doi: 10.1007/BF01279020
![]() |
[23] |
Z. M. Wu, Z. C. Xiong, Multivariate quasi-interpolation in Lp(Rd) with radial basis functions for scattered data, Int. J. Comput. Math., 87 (2010), 583–590. https://doi.org/10.1080/00207160802158702 doi: 10.1080/00207160802158702
![]() |
[24] |
L. Ling, A univariate quasi-multiquadric interpolation with better smoothness, Comput. Math. Appl., 48 (2004), 897–912. https://doi.org/10.1016/j.camwa.2003.05.014 doi: 10.1016/j.camwa.2003.05.014
![]() |
[25] |
R. H. Wang, M. Xu, Q. Fang, A kind of improved univariate multiquadric quasi-interpolation operators, Comput. Math. Appl., 59 (2010), 451–456. https://doi.org/10.1016/j.camwa.2009.06.023 doi: 10.1016/j.camwa.2009.06.023
![]() |
[26] |
R. Z. Feng, X. Zhou, A kind of multiquadric quasi-interpolation operator satisfying any degree polynomial reproduction property to scattered data, J. Comput. Appl. Math., 235 (2011), 1502–1514. https://doi.org/10.1016/j.cam.2010.08.037 doi: 10.1016/j.cam.2010.08.037
![]() |
[27] |
R. H. Wang, M. Xu, A kind of Bernoulli-type quasi-interpolation operator with univariate multiquadrics, Comput. Appl. Math., 29 (2010), 47–60. https://doi.org/10.1590/S1807-03022010000100004 doi: 10.1590/S1807-03022010000100004
![]() |
[28] |
R. F. Wu, H. L. Li, T. R. Wu, Univariate Lidstone-type multiquadric quasi-interpolants, Comput. Appl. Math., 39 (2020), 141. https://doi.org/10.1007/s40314-020-01159-x doi: 10.1007/s40314-020-01159-x
![]() |
[29] |
R. F. Wu, Abel-Goncharov Type Multiquadric Quasi-Interpolation Operators with Higher Approximation Order, J. Math., 2021 (2021), 1–12. https://doi.org/10.1155/2021/8874668 doi: 10.1155/2021/8874668
![]() |
[30] |
S. G. Zhang, C. G. Zhu, Q. J. Gao, Numerical Solution of High-Dimensional Shockwave Equations by Bivariate Multi-Quadric Quasi-Interpolation, Mathematics, 7 (2019), 734. https://doi.org/10.3390/math7080734 doi: 10.3390/math7080734
![]() |
[31] |
Z. M. Wu, R. Schaback, Shape preserving properties and convergence of univariate multiquadric quasi-interpolation, Acta. Math. Appl. Sin. Engl. Ser., 10 (1994), 441–446. https://doi.org/10.1007/BF02016334 doi: 10.1007/BF02016334
![]() |
[32] |
H. Y. Wu, Y. Duan, Multi-quadric quasi-interpolation method coupled with FDM for the Degasperis-Procesi equation, Appl. Math. Comput., 274 (2016), 83–92. https://doi.org/10.1016/j.amc.2015.10.044 doi: 10.1016/j.amc.2015.10.044
![]() |
[33] |
S. L. Zhang, H. Q. Yang, Y. Yang, A multiquadric quasi-interpolations method for CEV option pricing model, J. Comput. Appl. Math., 347 (2019), 1–11. https://doi.org/10.1016/j.cam.2018.03.046 doi: 10.1016/j.cam.2018.03.046
![]() |
[34] |
S. S. Li, Y. Duan, L. B. Li, On the meshless quasi-interpolation methods for solving 2D sine-Gordon euqations, Comput. Appl. Math., 41 (2022), 348. https://doi.org/10.1007/s40314-022-02054-3 doi: 10.1007/s40314-022-02054-3
![]() |
[35] | R. Jordan, Calculus of Finite Differences, New York: Chelsea Publishing Co, 1960. |
[36] |
F. A. Costabile, F. Dell'Accio, R. Luceri, Explicit polynomial expansions of regular real functions by means of even order Bernoulli polynomials and boundary values, J. Comput. Appl. Math., 176 (2005), 77–90. https://doi.org/10.1016/j.cam.2004.07.004 doi: 10.1016/j.cam.2004.07.004
![]() |
[37] | R. P. Agarwal, P. J. Y. Wong, Error Inequalities in Polynomial Interpolation and Their Applications, The Netherlands: Kluwer Academic Publishers, 1960. |
[38] | A. Sard, Linear Approximation, New York: AMS, Providence, RI, 1963. |
[39] |
D. D. Stancu, The remainder of certain linear approximation formulas in two variables, J. SIAM Numer. Anal. Ser. B, 1 (1964), 137–163. https://doi.org/10.1137/0701013 doi: 10.1137/0701013
![]() |
[40] | R. A. Devore, G. G. Lorentz, Constructive Approximation, New York: Springer-Verlag, 1993. |
[41] | Z. Ditzian, V. Totik, Moduli of Smoothness, New York: Springer-Verlag, Bernlin-Heidelberg, 1987. |
[42] |
R. J. Renka, A. K. Cline, A triangle-based C1 interpolation method, Rocky Mt. J. Math., 14 (1984), 223–237. https://doi.org/10.1216/RMJ-1984-14-1-223 doi: 10.1216/RMJ-1984-14-1-223
![]() |