On the Optimal Recovery of Integrals of Set-Valued Functions
We consider the problem of optimization of the approximate integration of set-valued functions from the class specified by a given majorant of their moduli of continuity performed by using the values of these functions at n fixed or free points of their domain.
Saved in:
| Date: | 2015 |
|---|---|
| Main Authors: | , , , , , |
| Format: | Article |
| Language: | English |
| Published: |
Institute of Mathematics, NAS of Ukraine
2015
|
| Online Access: | https://umj.imath.kiev.ua/index.php/umj/article/view/2055 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Journal Title: | Ukrains’kyi Matematychnyi Zhurnal |
| Download file: |  |
Institution
Ukrains’kyi Matematychnyi Zhurnal| _version_ | 1860507976545075200 |
|---|---|
| author | Babenko, V. F. Babenko, V. V. Polishchuk, M. V. Бабенко, В. Ф. Бабенко, В. В. Поліщук, М. В. |
| author_facet | Babenko, V. F. Babenko, V. V. Polishchuk, M. V. Бабенко, В. Ф. Бабенко, В. В. Поліщук, М. В. |
| author_sort | Babenko, V. F. |
| baseUrl_str | https://umj.imath.kiev.ua/index.php/umj/oai |
| collection | OJS |
| datestamp_date | 2019-12-05T09:49:43Z |
| description | We consider the problem of optimization of the approximate integration of set-valued functions from the class specified by a given majorant of their moduli of continuity performed by using the values of these functions at n fixed or free points of their domain. |
| first_indexed | 2026-03-24T02:17:52Z |
| format | Article |
| fulltext |
UDC 517.5
V. F. Babenko (Dnipropetrovsk Nat. Oles Honchar Univ.),
V. V. Babenko (Univ. of Utah, USA),
M. V. Polishchuk (Dnipropetrovsk Nat. Oles Honchar Univ.)
ON THE OPTIMAL RECOVERY OF INTEGRALS OF SET-VALUED FUNCTIONS
ПРО ОПТИМАЛЬНЕ ВIДНОВЛЕННЯ IНТЕГРАЛIВ
ВIД БАГАТОЗНАЧНИХ ФУНКЦIЙ
We consider the problem of optimization of the approximate integration of set-valued functions from the class defined by
a given majorant of their moduli of continuity performed by using the values of these functions at n fixed or free points of
their domain.
Розглядається задача оптимiзацiї наближеного iнтегрування на класах багатозначних функцiй, що мають задану
мажоранту модуля неперервностi. При цьому використовуються значення функцiй в n фiксованих або довiльних
точках областi визначення.
1. Introduction. By K(Rm) we denote the space of nonempty compact subsets of Rm. Let Kc(Rm)
be the set of convex elements of K(Rm). We consider below set-valued functions with nonempty
compact images, i.e., functions f : [0, 1]→ K(Rm).
Considerations of integration of set-valued functions go back to Minkowski and currently there
exist many different approaches to the definition of integrals of set-valued functions (see, e.g., [1 – 8]).
Integrals of such functions were found to be very applicable in many mathematical fields, especially
in mathematical economics, control theory, integral geometry, and statistics. One of the most useful
is Aumann integral [1] because this integral has many good properties. At the same time it is proved
in [2] that Riemann – Minkowski integral for any continuous and bounded set-valued function exists
and coincides with Aumann integral.
Theory of numerical integration is important part of approximation theory and numerical analysis
and works of many mathematicians were devoted to the problems of optimization of quadrature
formulas in various settings for the classes of real-valued functions. For surveys of obtained results
see, e.g., [9 – 12]. Estimates of deviation of Riemann sums and some other methods of approximate
calculation of integrals from the corresponding integrals of set-valued functions were considered in
the works [13 – 16]. Articles [17, 18] are devoted to the optimization of quadrature formulas on
classes of monotone with respect to inclusion convex-valued functions.
The goal of this paper is to consider the problems of optimization of approximate calculation of
Riemann – Minkowski integrals of set-valued functions from the class defined by given majorant of
their moduli of continuity (not necessary convex-valued functions) using values of the functions at n
fixed or n free points of their domain. Since Riemann – Minkowski integral is always a convex set, it
is not natural to use direct analogs of usual quadrature formulas. Instead, we consider these problems
from optimal recovery theory point of view.
Note that the theory of optimal recovery of functions, functionals, and operators has been incre-
mentally developed since mid 1960. Statements of the problems and surveys of obtained results can
be found in [19 – 27] and others.
Our paper is organized as follows. In Section 2 we present some necessary definitions and facts
from set-valued analysis. Statements of problems of optimal recovery of weighted integrals of set-
valued functions using exact values of the functions at n fixed or n free points of the domain are
c© V. F. BABENKO, V. V. BABENKO, M. V. POLISHCHUK, 2015
ISSN 1027-3190. Укр. мат. журн., 2015, т. 67, № 9 1163
1164 V. F. BABENKO, V. V. BABENKO, M. V. POLISHCHUK
presented in Section 3. Solution of the problem of optimal recovery of weighted integrals of set-valued
functions using exact values of the functions at n fixed points of the domain on the class of functions
having given majorant of moduli of continuity is presented in Section 4. The problem of optimal
recovery in the case when information is given at n free points of a domain is discussed in Section 5.
2. Preliminaries. In this section we present some definitions and facts from theory of set-valued
functions.
As usual, a linear combination of sets A,B ⊂ K(Rm) is defined by
λA+ µB = {λa+ µb : a ∈ A, b ∈ B} , λ, µ ∈ R.
Convex hull, denoted by coA, of a set A ⊂ K(Rm) is the set of all elements of the form
∑r
i=1
λiai,
where r ≥ 2, ai ∈ A, λi ∈ R, λi ≥ 0 for i = 1, . . . , r, and
∑r
i=1
λi = 1. Convex hull has the
following properties:
co (µA) = µ coA ∀µ ∈ R, A ⊂ K(Rm),
co (A+B) = coA+ coB ∀A,B ⊂ K(Rm).
If a = (a1, . . . , am) ∈ Rm, then ‖a‖ :=
√∑m
j=1
aj
2. For a point a ∈ Rm, and a set B ∈ K(Rm),
let d(a,B) := infb∈B ‖a− b‖ be the distance from the point a to the set B. For sets A,B ∈ K(Rm)
let
d(A,B) := sup
a∈A
d(a,B)
be the distance from the set A to the set B. Hausdorff metric δ in the space K(Rm) is defined as
follows. If A,B ∈ K(Rm), then
δ(A,B) := max{d(A,B), d(B,A)}.
Note that K(Rm) endowed with Hausdorff metric is a complete metric space.
Metric δ (A,B) has the following properties:
δ (λA, λB) = λδ (A,B) ∀λ > 0 ∀A,B ∈ K(Rm),
δ (A+B,C +D) ≤ δ (A,C) + δ (B,D) ∀A, B, C, D ∈ K(Rm),
δ(co A, co B) ≤ δ(A,B) ∀A,B ∈ K(Rm).
One can find proofs of all properties presented above in [3] and [28].
The Aumann’s integral of a globally bounded set-valued function f : [0, 1]→ K(Rm) is defined
as the set of all integrals of integrable selections of f [1]:
I(f) =
1∫
0
f(x)dx :=
1∫
0
φ(x)dx : φ(x) ∈ f(x) a. e., φ is integrable
.
The Riemann – Minkowski sum of f is defined in the following way. Let P = {x0, x1, . . . , xn},
0 = x0 < x1 < . . . < xn = 1, be some partition of the interval [0, 1]. We set ∆xi = xi − xi−1,
ISSN 1027-3190. Укр. мат. журн., 2015, т. 67, № 9
ON THE OPTIMAL RECOVERY OF INTEGRALS OF SET-VALUED FUNCTIONS 1165
λ(P ) = max{|∆xi| : i = 1, . . . , n}, and ξ = {ξ1, . . . , ξn}, ξi ∈ [xi−1, xi], i = 1, . . . , n. The
Riemann – Minkowski sum of f relative to the pair (P, ξ) is defined as
σ (f ; (P, ξ)) :=
n∑
i=1
∆xi · f(ξi).
We define the standard base λ(P )→ 0 in the set of all pairs (P, ξ) as follows [29]:
λ(P )→ 0 := {Bε}ε>0, Bε := {(P, ξ) : λ(P ) < ε}.
A function f is integrable in the Riemann – Minkowski sense if (see [2, 4]) there exists an element
I(f) ∈ K(Rm) such that
δ (σ (f ; (P, ξ)), I(f))→ 0 as λ(P )→ 0.
It is proved in [2] that Riemann – Minkowski integral for any continuous and bounded set-valued
function exists and coincides with Aumann integral.
By RM([0, 1],K(Rm)) we denote the set of functions which are integrable in the Riemann –
Minkowski sense. Note that bounded and continuous functions f : [0, 1] → K(Rm) belong to
RM([0, 1],K(Rm)), and the product P · f of a continuous real-valued function P and a function
f ∈ RM([0, 1],K(Rm)) belongs to RM([0, 1],K(Rm)). Below we denote by
∫ 1
0
f(x)dx the
Riemann – Minkowski integral for functions f ∈ RM([0, 1],K(Rm)).
Riemann – Minkowski integral has the following properties (see [3, 4, 28])
1∫
0
f(x)dx ∈ Kc(Rm) ∀f ∈ RM([0, 1],K(Rm)),
1∫
0
co(f(x))dx =
1∫
0
f(x)dx ∀f ∈ RM([0, 1],K(Rm)),
1∫
0
λf(x)dx = λ
1∫
0
f(x)dx ∀f ∈ RM([0, 1],K(Rm)) ∀λ ∈ R,
1∫
0
(f(x) + g(x))dx =
1∫
0
f(x)dx+
1∫
0
g(x)dx ∀f, g ∈ RM([0, 1],K(Rm)),
n∑
i=1
xi∫
xi−1
f(x) dx =
1∫
0
f(x) dx ∀f ∈ RM([0, 1],K(Rm)),
δ
1∫
0
f(x)dx,
1∫
0
g(x)dx
≤ 1∫
0
δ(f(x), g(x))dx ∀f, g ∈ RM([0, 1],K(Rm)).
ISSN 1027-3190. Укр. мат. журн., 2015, т. 67, № 9
1166 V. F. BABENKO, V. V. BABENKO, M. V. POLISHCHUK
3. Setting of the problems. LetM be some class of Riemann – Minkowski integrable functions
f : [0, 1]→ K(Rm), i.e.,M⊂ RM([0, 1],K(Rm)). Let continuous, nonnegative almost everywhere
function P : [0, 1] → R be given. Let also a set of points x = {x1, . . . , xn}, 0 ≤ x1 < x2 < . . .
. . . < xn ≤ 1, be given. We consider a problem of optimal recovery of the integral
1∫
0
P (x)f(x)dx
on the classM, using information f(x1), . . . , f(xn).
Arbitrary convex-valued mapping
Φ : K(Rm)× . . .×K(Rm︸ ︷︷ ︸
n times
)→ Kc(Rm)
is called a method of recovery of this integral.
The problem of finding the optimal method of recovery is formulated in the following way. Let
R(M, x,Φ) := sup
f∈M
δ
1∫
0
P (x)f(x)dx,Φ(f(x1), . . . , f(xn))
.
This value is called the error of a method Φ on the classM. Let also
R(M, x) := inf
Φ
R(M, x,Φ). (1)
The mapping Φ, that realizes inf
Φ
on the right-hand side of (1), is called optimal for the classM for
a fixed set of knots x.
Problem 1. Find the value
R(M, x) = inf
Φ
R(M, x,Φ),
and optimal method Φ.
Let now
Rn(M) := inf
#(x)=n
R(M, x), (2)
where #(x) is the number of elements in the set x.
This value is called the optimal error of recovery using n knots on the class M, and the set x∗
that realizes inf
#(x)=n
on the right-hand part of (2) is called an optimal set of knots.
Problem 2. Find the value Rn(M), optimal set of knots x∗, and optimal on the classM method
Φ that uses values f(x∗1), . . . , f(x∗n).
We solve Problems 1 and 2 for the following classes of set-valued functions. Given modulus of
continuity ω(t), we denote by Hω([0, 1],K(Rm)) the class of functions f : [0, 1] → K(Rm) such
that,
δ(f(x′), f(x′′)) ≤ ω(|x′ − x′′|) ∀ x′, x′′ ∈ [0, 1].
In Section 6 we consider the problems of optimal recovery of integrals on the class
Hω([0, 1],K(Rm)) using information with error.
ISSN 1027-3190. Укр. мат. журн., 2015, т. 67, № 9
ON THE OPTIMAL RECOVERY OF INTEGRALS OF SET-VALUED FUNCTIONS 1167
4. Solution of the Problem 1 for M = Hω([0, 1],K(Rm)). Let the set of knots x =
= {x1, . . . , xn} be given. We define
Πi(x) = {x ∈ [0, 1] : min
j=1,...,n
|x− xj | = |x− xi|},
c∗i = c∗i (P, x) =
∫
Πi(x)
P (x)dx, i = 1, . . . , n.
In particular if P (x) ≡ 1, then
c∗1 = c∗1(x) =
x1 + x2
2
,
c∗i = c∗i (x) =
xi+1 − xi−1
2
, if 1 < i < n,
c∗n = c∗n(x) = 1− xn−1 + xn
2
.
In addition, we define fω,x(x) := ω (mini=1,...,n |x− xi|).
Theorem 1. Let a modulus of continuity ω(t) and a set of points x = {x1, . . . , xn} be given.
The optimal method of the recovery of integral
∫ 1
0
P (x)f(x)dx on the class Hω([0, 1],K(Rm)),
using information f(x1), . . . , f(xn), is
Φ∗(f(x1), . . . , f(xn)) = co
(
n∑
k=1
c∗k(P, x)f(xk)
)
,
and optimal error of recovery is
R(Hω([0, 1],K(Rm)), x) =
1∫
0
P (x)fω,x(x)dx.
Proof. Using properties of Riemann – Minkowski integral, Hausdorff metric, and convex hull
presented in Section 2, we have for any f ∈ Hω([0, 1],K(Rm))
δ
1∫
0
P (x)f(x)dx, co
(
n∑
i=1
c∗i (P ;x)f(xi)
) =
= δ
n∑
i=1
∫
Πi(x)
P (x)cof(x)dx,
n∑
i=1
∫
Πi(x)
P (x)dx · cof(xi)
≤
≤
n∑
i=1
δ
∫
Πi(x)
P (x)cof(x)dx,
∫
Πi(x)
P (x)cof(xi)dx
≤
ISSN 1027-3190. Укр. мат. журн., 2015, т. 67, № 9
1168 V. F. BABENKO, V. V. BABENKO, M. V. POLISHCHUK
≤
n∑
i=1
∫
Πi(x)
P (x)δ(cof(x), cof(xi))dx ≤
n∑
i=1
∫
Πi(x)
P (x)δ(f(x), f(xi))dx ≤
≤
n∑
i=1
∫
Πi(x)
P (x)ω(|x− xi|)dx =
1∫
0
P (x)fω,x(x)dx.
Consequently,
R(Hω([0, 1],K(Rm)), x) ≤ R(Hω([0, 1],K(Rm)), x,Φ∗) ≤
1∫
0
P (x)fω,x(x)dx. (3)
We obtained the estimate from above for the value R(Hω([0, 1],K(Rm)), x). Next we obtain the
estimate from below.
We choose an arbitrary a ∈ Rm such that δ({a}, {θ}) = ‖a‖ = 1, where θ = (0, . . . , 0) ∈ Rm,
and define fω,x,a : [0, 1]→ K(Rm) with the help of the equality
fω,x,a(x) := fω,x(x) · {a}.
Note that fω,x,a(x) ∈ Hω([0, 1],K(Rm)), fω,x,a(xk) = {θ}, k = 1, . . . , n, and
1∫
0
P (x)fω,x,a(x)dx =
1∫
0
P (x)fω,x(x)dx · {a}.
For an arbitrary method of recovery Φ, we get
R(Hω([0, 1],K(Rm)), x,Φ) =
= sup
f∈Hω([0,1],K(Rm))
δ 1∫
0
P (x)f(x)dx,Φ (f(x1), . . . , f(xn))
≥
≥ max
δ
1∫
0
P (x)fω,x,a(x)dx,Φ ({θ}, . . . , {θ})
,
δ
− 1∫
0
P (x)fω,x,a(x)dx,Φ ({θ}, . . . , {θ})
=
= max
δ
1∫
0
P (x)fω,x(x)dx · {a},Φ ({θ}, . . . , {θ})
,
δ
− 1∫
0
P (x)fω,x(x)dx · {a},Φ ({θ}, . . . , {θ})
≥
ISSN 1027-3190. Укр. мат. журн., 2015, т. 67, № 9
ON THE OPTIMAL RECOVERY OF INTEGRALS OF SET-VALUED FUNCTIONS 1169
≥ 1
2
δ
1∫
0
P (x)fω,x(x)dx · {a},Φ ({θ}, . . . , {θ})
+
+δ
− 1∫
0
P (x)fω,x(x)dx · {a},Φ ({θ}, . . . , {θ})
≥
≥ 1
2
δ
1∫
0
P (x)fω,x(x)dx · {a},−
1∫
0
P (x)fω,x(x)dx · {a}
=
=
1
2
δ
2
1∫
0
P (x)fω,x(x)dx · {a}, {θ}
=
1∫
0
P (x)fω,x(x)dx · δ ({a}, {θ}) =
=
1∫
0
P (x)fω,x(x)dx.
Therefore, for an arbitrary method Φ
R(Hω([0, 1],K(Rm)), x,Φ) ≥
1∫
0
P (x)fω,x(x)dx. (4)
Comparing relations (3) and (4), we obtain the statement of Theorem 1.
Remark. Theorem 1 generalizes results of N. P. Korneichuk [30] and G. K. Lebed’ [31] for
real-valued functions. Multivariate analogs of their results were obtained by V. F. Babenko (see
[32 – 34]).
5. Optimal recovery of integrals using n free knots. It follows from Korneichuk and Lebed’
results, that Problem 2 will be solved for the class Hω([0, 1],K(Rm)) if we find the set of knots x0
that realizes
inf
x
1∫
0
P (x)fω,x(x)dx.
Then the optimal method is
Φ∗(f(x0
1), . . . , f(x0
n)) = co
(
n∑
k=1
c∗k(P, x
0)f(x0
k)
)
.
In addition,
Rn(Hω([0, 1],K(Rm))) =
1∫
0
P (x)fω,x0(x)dx.
Comparing this fact and Korneichuk’s result from [30], we obtain that in the case P (x) ≡ 1 the
following theorem holds.
ISSN 1027-3190. Укр. мат. журн., 2015, т. 67, № 9
1170 V. F. BABENKO, V. V. BABENKO, M. V. POLISHCHUK
Theorem 2. Let a modulus of continuity ω(t) and a number n ∈ N be given. Then
Rn(Hω([0, 1],K(Rm))) = 2n
1
2n∫
0
ω(t)dt,
optimal set of knots is x∗ = (x∗1, x
∗
2, . . . , x
∗
n) :=
(
1
2n
,
3
2n
, . . . ,
2n− 1
2n
)
, and the method
Φ(f(x∗1), . . . , f(x∗n)) = co
(
1
n
n∑
i=1
f
(
2i− 1
2n
))
is optimal on the class Hω([0, 1],K(Rm)) among all methods of recovery of the integral
∫ 1
0
f(x)dx
that use the information of the form f(x1), f(x2), . . . , f(xn).
In the case when P (x) is not identically equal to 1 we can not obtain the explicit expressions
for the optimal knots and the explicit value for Rn(Hω([0, 1],K(Rm))). However, using the results
from [35] we can obtain the exact asymptotics for this value (under some additional assumptions)
when n tends to ∞.
Let c > 0 be given. Let Ω(x) :=
∫ x
0
ω(t/2)dt, γc(x) := Ω−1(cΩ(x)), and
B(P, ω) := lim
n→∞
n∑
k=1
Ω−1
(
P
(
2k − 1
2n
)
Ω
(
1
n
))
.
Theorem 3. Let a modulus of continuity ω(x) be such that for all c > 0 the function γc(x)/x is
monotone in the right neighborhood of zero. Let also weight function P be continuous and positive
almost everywhere on [0, 1]. Then
lim sup
n→∞
Rn (Hω([0, 1],K(Rm)))
nΩ(B/n)
= 1.
Corollary 1. Let P (x) be the same as in the previous theorem. Let also ω(x) = xα, α ∈ (0, 1].
Then
Rn (Hω([0, 1],K(Rm))) =
(2n)−α
α+ 1
1∫
0
P (x)
1
1+αdx
α+1
+ o
(
1
nα
)
, n→∞.
1. Aumann R. J. Intgrals of set-valued functions// J. Math. Anal. and Appl. – 1965. – 12, № 1. – P. 1 – 12.
2. Polovinkin E. S. Riemannian integral of set-valued function // Lect. Notes Comput. Sci.: Optim. Techn., IFIP Techn.
Conf. Novosibirsk. – 1974. – 27. – P. 405 – 410.
3. Price G. B. The theory of integration // Trans. Amer. Math. Soc. – 1940.– 47. – P. 1 – 50.
4. Materon G. Random sets and Integral geometry. – New York etc.: John Wiley and Sons, 1975.
5. Dinghas A. Zum Minkowskishen Integralbegriff abgeschlossener Mengen // Math. Z. – 1956. – 66. – S. 173 – 188.
6. Debreu G. Integration of correspondence // Proc. First Berkelay Symp. Math. Statist. and Probab. – 1967. – P. 351 –
372.
7. Hukuhara M. Intégration des applications measurable dont la valeur est un compact convexe // Funkc. ekvacioj. –
1967. – 10. – P. 205 – 223.
ISSN 1027-3190. Укр. мат. журн., 2015, т. 67, № 9
ON THE OPTIMAL RECOVERY OF INTEGRALS OF SET-VALUED FUNCTIONS 1171
8. Artstein Z., Burns John A. Integration of compact set-valued functions // Pacif. J. Math. – 1975. – 58. – P. 297 – 307.
9. Nikol’skii S. M. Quadrature formulas (in Russian). – Moscow: Nauka, 1988.
10. Boyanov B. D. Optimal quadrature formulae // Rus. Math. Surv. – 2005. – 60(6). – P. 1035 – 1055.
11. Zhensykbaev A. A. Monosplines of minimal norm and the best quadrature formulae // Rus. Math. Surv. – 1981. – 36,
№ 4. – P. 121 – 180.
12. Korneichuk N. P., Ligun A. A., Babenko V. F. Extremal properties of polynomials and splines. – Nova Sci. Publ.,
1996.
13. Balaban E. I. On approximate evaluation of the Riemann integral of many-valued mapping // USSR Comput. Math.
and Math. Phys. – 1982. – 22, № 2. – P. 233 – 238.
14. Dontchev T., Farkhi E. Error estimates for discretized differential inclusions // Computing. – 1990. – 41. – P. 349 – 358.
15. Donchev T., Farkhi E. Moduli of smoothness of vector-valued functions of a real variable and applications // Numer.
Funct. Anal. Optim. – 1990. – 11(586). – P. 497 – 509.
16. Baier R., Lempio F. Computing Aumann’s integral // Modeling Techniques for Uncertain Systems: Progress in
Systems and Control Theory / Eds A. B. Kurzhanski, V. M. Vehov. – Basel: Birkhäuser, 1994. – 18. – P. 71 – 92.
17. Babenko V. F., Babenko V. V. Optimization of approximate integration of set-valued functions monotone with respect
to inclusion // Ukr. Math. J. – 2011. – 63, № 2. – P. 177 – 186.
18. Babenko V. V. Optimization of interval formulas for approximate integration of set-valued functions monotone with
respect to inclusion // Ukr. Math. J. – 2012. – 63, № 11. – P. 1781 – 1788.
19. Smolyak S. A. On optimal recovery of functions and functionals of them (in Russian): Diss. . . . cand. fiz.-mat. nauk.
– M., 1965.
20. Bakhvalov N. S. On the optimality of linear methods for operator approximation in convex classes of functions //
Comput. Math. and Math. Phys. USSR. – 1971. – 11, № 4. – P. 244 – 249.
21. Michelli C. A., Rivlin T. J. Lectures on optimal recovery // Lect. Notes Math. Numer. Anal. – Berlin: Springer-Verlag,
1984. – P. 21 – 93.
22. Magaril-Il’yaev G. G., Osipenko K. Yu. Optimal recovery of functionals based on inaccurate data // Math. Notes
Acad. Sci. USSR. – 1991. – 50, № 6. – P. 1274 – 1279.
23. Osipenko K. Yu. Optimal recovery of analytic functions. – New York: Nova Sci. Publ., 2000.
24. Traub J., Woźniakowski H. A general theory of optimal algorithms // ACM Monogr. Ser. – New York etc.: Acad.
Press, 1980. – Vol. 15.
25. Traub J., Wasilkowski G., Woźniakowski H. Information-based complexity // Comput. Sci. and Sci. Computing. –
Boston, MA: Acad. Press, Inc., 1988.
26. Arestov V. V. Approximation of unbounded operators by bounded operators and related extremal problems // Rus.
Math. Surv. – 1996. – 51, № 6. – P. 1093 – 1126.
27. Zhensykbaev A. A. Problems of recovery of operators. – Moscow; Izhevsk: Inst. Comput. Stad., 2003.
28. Polovinkin E. S., Balashov M. V. Elements of convex and strongly convex analysis (in Russian). – Moscow: Fizmatlit,
2004.
29. Zorich V. A. Mathematical analysis. – Berlin; Heidelberg: Springer-Verlag, 2009. – Vol. 1.
30. Korneichuk N. P. Best cubature formulas for some classes of functions of many variables // Math. Notes Acad. Sci.
USSR. – 1968. – 3, № 5. – P. 360 – 367.
31. Lebed’ G. K. Quadrature formulas with minimum error for certain classes of functions // Math. Notes Acad. Sci.
USSR. – 1968. – 3, № 5. – P. 368 – 373.
32. Babenko V. F. Asymptotically sharp bounds for the remainder for the best quadrature formulas for several classes of
functions // Math. Notes Acad. Sci. USSR. – 1976. – 19, № 3. – P. 187 – 193.
33. Babenko V. F. Faithful asymptotics of remainders optimal for some glasses of functions with cubic weight formulas
// Math. Notes Acad. Sci. USSR. – 1976. – 20, № 4. – P. 887 – 890.
34. Babenko V. F. On the optimal error bound for cubature formulae on certain classes of continuous functions // Anal.
Math. – 1977. – 3. – P. 3 – 9.
35. Babenko V. F. On optimization of weight quadrature formulas // Ukr. Math. J. – 1995. – 47, № 8. – P. 1157 – 1168.
Received 29.05.14
ISSN 1027-3190. Укр. мат. журн., 2015, т. 67, № 9
|
| id | umjimathkievua-article-2055 |
| institution | Ukrains’kyi Matematychnyi Zhurnal |
| keywords_txt_mv | keywords |
| language | English |
| last_indexed | 2026-03-24T02:17:52Z |
| publishDate | 2015 |
| publisher | Institute of Mathematics, NAS of Ukraine |
| record_format | ojs |
| resource_txt_mv | umjimathkievua/5d/c1ad27bdea55fc1af4d079ce0aeea85d.pdf |
| spelling | umjimathkievua-article-20552019-12-05T09:49:43Z On the Optimal Recovery of Integrals of Set-Valued Functions Про оптимальне відновлення інтегралів від багатозначних функцій Babenko, V. F. Babenko, V. V. Polishchuk, M. V. Бабенко, В. Ф. Бабенко, В. В. Поліщук, М. В. We consider the problem of optimization of the approximate integration of set-valued functions from the class specified by a given majorant of their moduli of continuity performed by using the values of these functions at n fixed or free points of their domain. Розглядається задача оптимізації наближеного інтегрування на класах багатозначних функцій, що мають задану мажоранту модуля неперервності. При цьому використовуються значення функцій в n фіксованих або довільних точках області визначення. Institute of Mathematics, NAS of Ukraine 2015-09-25 Article Article application/pdf https://umj.imath.kiev.ua/index.php/umj/article/view/2055 Ukrains’kyi Matematychnyi Zhurnal; Vol. 67 No. 9 (2015); 1163-1171 Український математичний журнал; Том 67 № 9 (2015); 1163-1171 1027-3190 en https://umj.imath.kiev.ua/index.php/umj/article/view/2055/1126 https://umj.imath.kiev.ua/index.php/umj/article/view/2055/1127 Copyright (c) 2015 Babenko V. F.; Babenko V. V.; Polishchuk M. V. |
| spellingShingle | Babenko, V. F. Babenko, V. V. Polishchuk, M. V. Бабенко, В. Ф. Бабенко, В. В. Поліщук, М. В. On the Optimal Recovery of Integrals of Set-Valued Functions |
| title | On the Optimal Recovery of Integrals of Set-Valued Functions |
| title_alt | Про оптимальне відновлення інтегралів від багатозначних функцій |
| title_full | On the Optimal Recovery of Integrals of Set-Valued Functions |
| title_fullStr | On the Optimal Recovery of Integrals of Set-Valued Functions |
| title_full_unstemmed | On the Optimal Recovery of Integrals of Set-Valued Functions |
| title_short | On the Optimal Recovery of Integrals of Set-Valued Functions |
| title_sort | on the optimal recovery of integrals of set-valued functions |
| url | https://umj.imath.kiev.ua/index.php/umj/article/view/2055 |
| work_keys_str_mv | AT babenkovf ontheoptimalrecoveryofintegralsofsetvaluedfunctions AT babenkovv ontheoptimalrecoveryofintegralsofsetvaluedfunctions AT polishchukmv ontheoptimalrecoveryofintegralsofsetvaluedfunctions AT babenkovf ontheoptimalrecoveryofintegralsofsetvaluedfunctions AT babenkovv ontheoptimalrecoveryofintegralsofsetvaluedfunctions AT políŝukmv ontheoptimalrecoveryofintegralsofsetvaluedfunctions AT babenkovf prooptimalʹnevídnovlennâíntegralívvídbagatoznačnihfunkcíj AT babenkovv prooptimalʹnevídnovlennâíntegralívvídbagatoznačnihfunkcíj AT polishchukmv prooptimalʹnevídnovlennâíntegralívvídbagatoznačnihfunkcíj AT babenkovf prooptimalʹnevídnovlennâíntegralívvídbagatoznačnihfunkcíj AT babenkovv prooptimalʹnevídnovlennâíntegralívvídbagatoznačnihfunkcíj AT políŝukmv prooptimalʹnevídnovlennâíntegralívvídbagatoznačnihfunkcíj |