Uniform approximation by positive operators on infinite intervals

В работе устанавливае тся оценка (*) |L _n (???| ≦Kω _? (?;α _n) для положительных оп ераторов, определенн ых на конечном или бесконе чном интервале (a,b), гдеL _n(1,χ)≡1,L _n((t?χ)²;χ)≦K? ²(χ)α _n ² (x∈(a,b)) ;и \(\omega _\varphi (f;\delta ) = \mathop {\sup }\limits_{0 \leqq h \leqq \delta ,x \pm h\varphi (x) \in (a,b)} \left| {f(x - h\varphi (x)) - 2f(x) + f(x + h\varphi (x))} \right|\) модуль гладкости?, св язанный с ? (функция? удовлетворяет некот орым условиям регуля рности). С помощью (*) для некотор ых {L _n } получена характеристика тех ф ункций?, для которыхL _n (?)??=o(1) равном ерно на (a, b). Наконец, рассматриваются слу чай насыщения и случай так называем ой неоптимальной апп роксимации. Результаты применяю тся к операторам Саса —Миракяна, Баскакова, Мейер-Кëни га и Целлера, гамма и бета операторам, а также к н екоторым операторам типа свер тки.     

Variational perturbations of the linear-quadratic problem

A sequence of perturbations $$\begin{gathered} \dot x = A_n x + B_n u, \parallel u\parallel _{L^2 } \leqslant 1, (P_n ) \hfill \\ x\left( o \right) = x^o , n = 0, 1, 2, 3,..., \hfill \\ \end{gathered} $$ is given of the linear-quadratic optimal control problem consisting of minimizing $$\int_0^1 {((u - \tilde u)^T (u - \tilde u) + (x - \tilde x)^T (x - \tilde x))dt,} $$ subject to (P₀). We assume that {A _n} bounded inL ¹ and {B _n} is bounded inL ². Then, a necessary and sufficient condition so that, for every?, \(\tilde u\) , \(\tilde x\) ∈L ², and for everyx ⁰, the optimal control for (P_n) converges strongly inL ² to the optimal control for (P₀) and the optimal state converges uniformly is thatA _n →A ₀ weakly inL ¹ andB _n →B ₀ strongly inL ².     

Convergence of biorthogonal series in the system of contractions and translations of functions in the spaces L p [0, 1]

We obtain conditions for the convergence in the spaces L ^p [0, 1], 1 ≤ p < ∞, of biorthogonal series of the form $$ f = \sum\limits_{n = 0}^\infty {(f,\psi _n )\phi _n } $$ in the system {? _n } _n≥0 of contractions and translations of a function ?. The proposed conditions are stated with regard to the fact that the functions belong to the space $ \mathfrak{L}^p $ of absolutely bundleconvergent Fourier-Haar series with norm $$ \left\| f \right\|_p^ * = \left| {f,\chi _0 } \right| + \sum\limits_{k = 0}^\infty {2^{k({1 \mathord{\left/ {\vphantom {1 2}} \right. \kern-\nulldelimiterspace} 2} - {1 \mathord{\left/ {\vphantom {1 p}} \right. \kern-\nulldelimiterspace} p})} } \left( {\sum\limits_{n = 2^k }^{2^{k + 1} - 1} {\left| {f,\chi _n } \right|^p } } \right)^{{1 \mathord{\left/ {\vphantom {1 p}} \right. \kern-\nulldelimiterspace} p}} , $$ where (f,χ _n ), n = 0, 1, ..., are the Fourier coefficients of a function f ? L ^p [0, 1] in the Haar system {χ _n } _n≥0. In particular, we present conditions for the system {? _n } _n≥0 of contractions and translations of a function ? to be a basis for the spaces L ^p [0, 1] and $ \mathfrak{L}^p $ .     

On the basis property of a trigonometric system arising in the Frankl problem

We consider the function system {cos4nθ} _n=0 ^∞ , {sin(4n ? 1)θ} _n=1 ^∞ , which arises in the Frankl problem in the theory of elliptic-hyperbolic equations. We show that this system is a Riesz basis in the space L ₂(0, π/2) and construct the biorthogonal system.     

Convergence of weighted sums of tight random elements

Convergence of weighted sums of tight random elements {V_n} (in a separable Banach space) which have zero expected values and uniformly bounded rth moments (r > 1) is obtained. In particular, if {a_nk} is a Toeplitz sequence of real numbers, then | Σ_k=1^∞a_nkf(V_k)| → 0 in probability for each continuous linear functional f if and only if 6Σ_k=1^∞a_nkV_k 6→ 0 in probability. When the random elements are independent and max_{1≤k≤n | ank | = $\text{O}$(n^?8)} for some $\text{0 <}\text{1}\text{s}\text{< r ? 1}$, then |Σ_k=1^∞a_nkV_k 6→ 0 with probability 1. These results yield laws of large numbers without assuming geometric conditions on the Banach space. Finally, these results can be extended to random elements in certain Fréchet spaces.     

Bounds for singular integrals associated with a class of hypersurfaces

For the hypersurface Γ=(y,γ(y)), the singular integral operator along Γ is defined by. $$Tf(x,x_n ) = P.V.\int_{\mathbb{R}^n } {, f(x - y,x_n ) - } \gamma (y))_{\left| y \right|^{n - 1} }^{\Omega (v)} dy$$ where Σ is homogeneous of order 0, $ \int_{\Sigma _{n \lambda } } {\Omega (y')dy'} = 0 $ . For a certain class of hypersurfaces, T is shown to be bounded on L^p(Rⁿ) provided Ω∈L _α ¹ (Σ _n?2),P>1.     

Kernels of sequences of complex numbers and their regular transformations

It is proved that , where U(a, r) is the ball of radius r with center at the pointa, is the smallest closed convex set containing the kernel of any sequence {y_n} obtained from the sequence {x_n} by means of a regular transformation (c_nk) satisfying the condition , where x, x_n, c_nk (n, k=1, 2,...) are complex numbers.Traslated from Matematicheskie Zametki, Vol. 22, No. 6, pp. 815–823, December, 1977.     

Convergence of weighted sums of independent random variables

Let {X_k} be a sequence of i.i.d. random variables with d.f. F(x). In the first part of the paper the weak convergence of the d.f.'s F_n(x) of sums is studied, where 0<α≤2, a_nk>0, 1≤k≤m_n, and, as n→∞, bothmax _1≤k≤mn_{a nk}→0 and . It is shown that such convergence, with suitably chosen A_n's and necessarily stable limit laws, holds for all such arrays {α_nk} provided it holds for the special case α_nk=1/n, 1≤k≤n. Necessary and sufficient conditions for such convergence are classical. Conditions are given for the convergence of the moments of the sequence {F_n(x)}, as well as for its convergence in mean. The second part of the paper deals with the almost sure convergence of sums , where a_n≠0, b_n>0, andmax _1≤k≤n a_k/b_n→0. The strong law is said to hold if there are constants A_n for which S_n→0 almost surely. Let N(0)=0 and N(x) equal the number of n≥1 for which b_n/|a_n|<x if x>0. The main result is as follows. If the strong law holds,EN (|X₁|)<∞. If for some 0<p≤2, then the strong law holds with if 1≤p≤2 and A_n=0 if 0<p<1. This extends the results of Heyde and of Jamison, Orey, and Pruitt. The strong law is shown to hold under various conditions imposed on F(x), the coefficients a_n and b_n, and the function N(x). Proceedings of the Seminar on Stability Problems for Stochastic Models, Moscow, 1993.     

L p-approximation by Kantorovič operators

Оператор Канторович а дляf∈L _p(I), I=[0,1], определяе тся соотношением $$P_n (f,x) = (n + 1)\sum\limits_{k = 0}^n {\left( {\begin{array}{*{20}c} n \\ k \\ \end{array} } \right)} x^k (1 - x)^{n - 1} \int\limits_{I_k } {f(t)dt,} $$ гдеI _k=[k/(n}+1),(k+1)/(n+ 1)],n∈N. Доказывается, что есл ир>1 иf∈W _p ² (I), т.е.f абсол ютно непрерывна наI иf″∈L _p(I), то $$\left\| {P_n f - f} \right\|_p = O(n^{ - 1} ).$$ Далее, установлено, чт о еслиf∈L _p(I),p>1 и ∥P _n f-f∥_р=О(n ^?1), тоf∈S, гдеS={f∶f аб-солютно непрерывна наI, x(1?x)f′(x)=∝ ₀ ^x h(t)dt, гдеh∈L _p(I) и ∝ ₀ ¹ h(t)dt=0}. Если жеf∈L_p(I),p>1, то из условия ∥P _n(f)?f∥_pL=o(n^?1) вытекает, чтоf постоянна почти всюду.     

On a property of complete orthonormal systems

In this paper we establish a necessary and sufficient condition for orthonormal systems, subject to which there exist rearranged series ∑_σα_n?_n(x) converging almost everywhere to functions h(x) ¯? L₂[0, 1]. In particular, we show, for an arbitrary complete orthonormal system, that such series exist.     

Local properties of functions and approximation by trigonometric polynomials

Suppose Φ_{p, E} (p>0 an integer, E ?[0, 2π]) is a family of positive nondecreasing functions? _x(t) (t>0, x∈ E) such that? _x(nt)≤n^P ? _x(t) (n=0,1,...), t_n is a trigonometric polynomial of order at most n, and Δ _h ^l (f, x) (l>0 an integer) is the finite difference of orderl with step h of the functionf.THEOREM. Supposef (x) is a function which is measurable, finite almost everywhere on [0, 2π], and integrable in some neighborhood of each point xε E,? _X εΦ_p,E and $$\overline {\mathop {\lim }\limits_{\delta \to \infty } } |(2\delta )^{ - 1} \smallint _{ - \delta }^\delta \Delta _u^l (f,x)du|\varphi _x^{ - 1} (\delta ) \leqslant C(x)< \infty (x \in E).$$ . Then there exists a sequence {t _n } _n=1 ^∞ which converges tof (x) almost everywhere, such that for x ε E $$\overline {\mathop {\lim }\limits_{n \to \infty } } |f(x) - l_n (x)|\varphi _x^{ - 1} (l/n) \leqslant AC(x),$$ where A depends on p andl.     

Limiting distributions of two random sequences

For fixed p (0 ≤ p ≤ 1), let {L₀, R₀} = {0, 1} and X₁ be a uniform random variable over {L₀, R₀}. With probability p let {L₁, R₁} = {L₀, X₁} or = {X₁, R₀} according as $\text{X}{}_1\text{≥}\text{1}\text{2}\text{(L}{}_0\text{+ R}{}_0\text{)}\text{or}\text{<}\text{1}\text{2}\text{(L}{}_0\text{+ R}{}_0\text{)}$; with probability 1 ? p let {L₁, R₁} = {X₁, R₀} or = {L₀, X₁} according as $\text{X}{}_1\text{≥}\text{1}\text{2}\text{(L}{}_0\text{+ R}{}_0\text{)}\text{or}\text{<}\text{1}\text{2}\text{(L}{}_0\text{+ R}{}_0\text{)}$, and let X₂ be a uniform random variable over {L₁, R₁}. For n ≥ 2, with probability p let {L_n, R_n} = {L_{n ? 1}, X_n} or = {X_n, R_{n ? 1}} according as $\text{X}{}_{\mathrm{n}}\text{≥}\text{1}\text{2}\text{(L}{}_{\mathrm{n\; ?\; 1}}\text{+ R}{}_{\mathrm{n\; ?\; 1}}\text{)}\text{or}\text{<}\text{1}\text{2}\text{(L}{}_{\mathrm{n\; ?\; 1}}\text{+ R}{}_{\mathrm{n\; ?\; 1}}\text{)}$, with probability 1 ? p let {L_n, R_n} = {X_n, R_{n ? 1}} or = {L_{n ? 1}, X_n} according as $\text{X}{}_{\mathrm{n}}\text{≥}\text{1}\text{2}\text{(L}{}_{\mathrm{n\; ?\; 1}}\text{+ R}{}_{\mathrm{n\; ?\; 1}}\text{)}\text{or}\text{<}\text{1}\text{2}\text{(L}{}_{\mathrm{n\; ?\; 1}}\text{+ R}{}_{\mathrm{n\; ?\; 1}}\text{)}$, and let X_{n + 1} be a uniform random variable over {L_n, R_n}. By this iterated procedure, a random sequence {X_n}_{n ≥ 1} is constructed, and it is easy to see that X_n converges to a random variable Y_p (say) almost surely as n → ∞. Then what is the distribution of Y_p? It is shown that the Beta, (2, 2) distribution is the distribution of Y₁; that is, the probability density function of Y₁ is g(y) = 6y(1 ? y) I_0,1(y). It is also shown that the distribution of Y₀ is not a known distribution but has some interesting properties (convexity and differentiability).     

A multipoint method of third order

LetF be a mapping of the Banach spaceX into itself. A convergence theorem for the iterative solution ofF(x)=0 is proved for the multipoint algorithmx _n+1=x _n ?ø(x _n ), where $$\phi (x) = F\prime_x^{ - 1} \left[ {F(x) + F\lgroup {x - F\prime_x^{ - 1} F(x)} \rgroup} \right]$$ andF′x is the Frechet derivative ofF. The theorem guarantees that, under appropriate conditions onF, the multipoint sequence {x _n } generated by ø converges cubically to a zero ofF. The algorithm is applied to the nonlinear Chandrasekhar integral equation $$\frac{1}{2}\omega _0 x(t)\int_0^1 {\frac{{tx(s)}}{{s + t}}ds - x(t) + 1 = 0}$$ where ω₀>0. A discretization of the equations of iteration is discussed, and some numerical results are given.     

On L 1-convergence of Fourier series of complex valued functions under the GM7 condition

Let f??L _2?? be a real-valued even function with its Fourier series $\frac {a_{0}}{2}+\sum\limits_{n=1}^{\infty}{a_{n}}\cos nx$ , and let S _n (f,x) be the nth partial sum of the Fourier series, n?R1. The classical result says that if the nonnegative sequence {a _n } is decreasing and $\lim\limits_{n\to\infty} a_{n} =0$ , then $\lim\limits_{n\to\infty} \|{f-S_{n}(f)}\|_{L}=0$ if and only if $\lim\limits_{n\to\infty} a_{n}\log n=0$ . Later, the monotonicity condition set on {a _n } is essentially generalized to MVBV (Mean Value Bounded Variation) condition. Very recently, Kórus further generalized the condition in the classical result to the so-called GM₇ condition in real space. In this paper, we give a complete generalization to the complex space.     

On the Riesz basis property of the eigen- and associated functions of periodic and antiperiodic Sturm-Liouville problems

The paper deals with the Sturm-Liouville operator $$ Ly = - y'' + q(x)y, x \in [0,1], $$ generated in the space L ₂ = L ₂[0, 1] by periodic or antiperiodic boundary conditions. Several theorems on the Riesz basis property of the root functions of the operator L are proved. One of the main results is the following. Let q belong to the Sobolev spaceW ₁ ^p [0, 1] for some integer p ≥ 0 and satisfy the conditions q ^(k)(0) = q ^(k)(1) = 0 for 0 ≤ k ≤ s ? 1, where s ≤ p. Let the functions Q and S be defined by the equalities $$ Q(x) = \int_0^x {q(t)dt, S(x) = Q^2 (x)} $$ and let q _n , Q _n , and S _n be the Fourier coefficients of q, Q, and S with respect to the trigonometric system $ \{ e^{2\pi inx} \} _{ - \infty }^\infty $ . Assume that the sequence q _2n ? S _2n + 2Q ₀ Q _2n decreases not faster than the powers n ^?s?2. Then the system of eigenfunctions and associated functions of the operator L generated by periodic boundary conditions forms a Riesz basis in the space L ₂[0, 1] (provided that the eigenfunctions are normalized) if and only if the condition $$ q_{2n} - s_{2n} + 2Q_0 Q_{2n} \asymp q_{ - 2n} - s_{2n} + 2Q_0 Q_{ - 2n} , n > 1, $$ holds.     

Approximation by antiderivatives

Let Ω ?C be an open set with simply connected components and suppose that the functionφ is holomorphic on Ω. We prove the existence of a sequence {φ ^(?n)} ofn-fold antiderivatives (i.e., we haveφ ⁽⁰⁾(z)∶=φ(z) andφ ^(?n)(z)=dφ ^(?n?1)(z)/dz for alln ∈ N₀ and z ∈ Ω) such that the following properties hold:

1. For any compact setB ?Ω with connected complement and any functionf that is continuous onB and holomorphic in its interior, there exists a sequence {n _k} such that {φ^?nk} converges tof uniformly onB.
2. For any open setU ?Ω with simply connected components and any functionf that is holomorphic onU, there exists a sequence {m _k} such that {φ^?mk} converges tof compactly onU.
3. For any measurable setE ?Ω and any functionf that is measurable onE, there exists a sequence {p _k} such that {φ ^(-Pk)} converges tof almost everywhere onE.

     

Constructive techniques for approximating collocation linear systems

A high order discretization by spectral collocation methods of the elliptic problem $$\begin{gathered} \left\{ \begin{gathered} A_{A,b} u \equiv - \nabla [A(x)\nabla u(x)] + b(x)u(x) = c(x){\text{ on}}\;\Omega \; = ( - 1,1)^2 \hfill \\ u\left| {\partial \Omega \equiv 0,} \right. \hfill \\ \end{gathered} \right. \hfill \\ \hfill \\ \end{gathered} $$ is considered where A(x)=a(x)I ₂, x=(x ^[1],x ^[2]) and I ₂ denotes the 2×2 identity matrix, giving rise to a sequence of dense linear systems that are optimally preconditioned by using the sparse Finite Difference (FD) matrix-sequence {A _n } _n over the nonuniform grid sequence defined via the collocation points [11]. Here we propose a preconditioning strategy for {A _n } _n based on the “approximate factorization” idea. More specifically, the preconditioning sequence {P _n } _n is constructed by using two basic structures: a FD discretization of (1) with A(x)=I ₂ over the collocation points, which is interpreted as a FD discretization over an equidistant grid of a suitable separable problem, and a diagonal matrix which adds the informative content expressed by the weight function a(x). The main result is the proof that the sequence {P _n ^?1 A _n } _n is spectrally clustered at unity so that the solution of the nonseparable problem (1) is reduced to the solution of a separable one, this being computationally more attractive [2,3]. Several numerical experiments confirm the goodness of the discussed proposal.     

Limit distributions for the maxima of stationary Gaussian processes

Let {X_n} be a stationary Gaussian sequence with E{X₀} = 0, {X²₀} = 1 and E{X₀X_n} = rⁿ_n Let c_n = (2ln n)^{$\text{built1}\text{2}$}, b_n = c_n? $\text{1}\text{2}$c^-1_n ln(4π ln n), and set M_n = max_{0 ?k?n}X_k. A classical result for independent normal random variables is that

$\text{P[c}{}_{\mathrm{n}}\text{(M}{}_{\mathrm{n}}\text{?b}{}_{\mathrm{n}}\text{)?x]→exp[-e}{}^{\mathrm{-x}}\text{] as n → ∞ for all x.}$

Berman has shown that (1) applies as well to dependent sequences provided r_nlnn = o(1). Suppose now that {r_n} is a convex correlation sequence satisfying r_n = o(1), (r_nlnn)^-1 is monotone for large n and o(1). Then

$\text{P[r}{}_{\mathrm{n}}{}^{\mathrm{<mtext>-1</mtext><mtext>2</mtext>}}\text{(M}{}_{\mathrm{n}}\text{? (1?r}{}_{\mathrm{n}}\text{)}{}^{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}\text{b}{}_{\mathrm{n}}\text{)?x] → Ф(x)}$

for all x, where Ф is the normal distribution function. While the normal can thus be viewed as a second natural limit distribution for {M_n}, there are others. In particular, the limit distribution is given below when r_n is (sufficiently close to) γ/ln n. We further exhibit a collection of limit distributions which can arise when r_n decays to zero in a nonsmooth manner. Continuous parameter Gaussian processes are also considered. A modified version of (1) has been given by Pickands for some continuous processes which possess sufficient asymptotic independence properties. Under a weaker form of asymptotic independence, we obtain a version of (2).     

Continuous additive functions and difference equations of infinite order

Пустьq∈(1, 2) иL=(q?1)^?1. Дляz∈[0,L] обозначимδ(z) функцию, для которойδ(z)=1, еслиz≧1/q иδ(z)=0, еслиz<1/q. Пустьy(z) определяется из урав ненияz= =δ(z)q ^?1+y(z)q ^?1, и регулярное представление \(\mathop \Sigma \limits_{n = 1}^\infty \varepsilon _n \left( x \right)q^{ - n} \) аргументах определя ется из следующих соотношен ий: $$x = x_0 , \varepsilon _n \left( x \right) = \delta \left( {x_n } \right), x_{n + 1} = y\left( {x_n } \right).$$ ФункцияF: [0,L]→C называе тся аддитивной, если о на представляется в вид е $$F\left( x \right) = \mathop \Sigma \limits_{n = 1}^\infty \varepsilon _n \left( x \right)a_n ,$$ где ε ¦a _n ¦<∞. «Бесконеч ное» представление 1=εl _i q ^?1 числа 1 определяется с ледующим образом: еслие _n (1)=1 для б есконечно многихп, т оl _n =ε _n (1) (n=1, 2, ...); если ? максим альный индекс, для которогоε _s (1)=1, то $$l_{ks + 1} = \left\{ \begin{gathered} \varepsilon _i \left( 1 \right) \left( {k = 0, 1, 2, ...; i = 1, ..., s - 1} \right) \hfill \\ 0 \left( {i = 0; k = 1, 2, ...} \right). \hfill \\ \end{gathered} \right.$$ В более ранней работе, опубликованной в это м журнале, авторы доказали, что а ддитивная функция является неп рерывной на отрезке [0,L] тогда и только тогда, когда ра венство $$a_n = \mathop \Sigma \limits_{i = 1}^\infty l_i a_{n + 1} $$ выполняется для всехn∈N. В настоящей работе ра ссматриваются непре рывные функции для которых в ыполняются дополнительные усло вия видаa _n =O(q ^??n ) (0a _n ≧0. Анализируются их свя зи с корнями функцииG(z)=1 +ε l _i z ⁱ . Доказы вается, что непрерывн ая аддитивная функция и ли вляется линейной, или нигде не дифференцир уема на отрезке [0,L].