The evaluation of Tornheim double sums. Part 2

We provide an explicit formula for the Tornheim double series T(a,0,c) in terms of an integral involving the Hurwitz zeta function. For integer values of the parameters, a=m, c=n, we show that in the most interesting case of even weight N:=m+n the Tornheim sum T(m,0,n) can be expressed in terms of zeta values and the family of integrals

Estimation of a Normal Covariance Matrix with Incomplete Data under Stein′s Loss

Suppose that we have (n − a) independent observations from N_p(0, Σ) and that, in addition, we have a independent observations available on the last (p − c) coordinates. Assuming that both observations are independent, we consider the problem of estimating Σ under the Stein′s loss function, and show that some estimators invariant under the permutation of the last (p − c) coordinates as well as under those of the first c coordinates are better than the minimax estimators of Eaten. The estimators considered outperform the maximum likelihood estimator (MLE) under the Stein′s loss function as well. The method involved here is computation of an unbiased estimate of the risk of an invariant estimator considered in this article. In addition we discuss its application to the problem of estimating a covariance matrix in a GMANOVA model since the estimation problem of the covariance matrix with extra data can be regarded as its canonical form.     

On the moments of some first passage times for sums of dependent random variables

Let S_n,n = 1, 2, …, denote the partial sums of integrable random variables. No assumptions about independence are made. Conditions for the finiteness of the moments of the first passage times N(c) = min {n: S_n>ca(n)}, where c ≥ 0and a(y) is a positive continuous function on [0, ∞), such that a(y) = o(y)as y → ∞, are given. With the further assumption that a(y) = yP,0 ≤ p < 1, a law of large numbers and the asymptotic behaviour of the moments when c → ∞ are obtained. The corresponding stopped sums are also studied.     

Finite metric spaces needing high dimension for lipschitz embeddings in banach spaces

We construct a sequence of metric spaces (M _n) with cardM _n=3n satisfying that for everyc<2, there exists a real numbera(c)>0 such that, if the Lipschitz distance fromM _n to a subset of a Banach spaceE is less thanc, then dim(E) ≥a(c)n. We also prove several results about embeddings of metric spaces whose non-zero distance values are in the interval [1,2].     

Eventually areally meanp-valent functions

Theorems concerning areally meanp-valent functions are extended to eventually areally meanp-valent functions. In particular, suppose is eventually areally meanp-valent in the unit disc,b, c are positive integers,a≧max {p−1, 0}. If |a _n|≦Cn ^α for alln=bm+c,m=1, 2, …, then |a _n|≦C′n ^α for alln. This is a marked extension of results due to Goluzin and to Hayman.     

Arithmetic of semigroups of series in multiplicative systems

We study the arithmetic of a semigroup M_P\mathcal{M}_{\mathcal{P}} of functions with operation of multiplication representable in the form f(x) = ?_{n = 0}^￥ a_nc_n(x)    ( a_n 3 0,?_{n = 0}^￥ a_n = 1 ) f(x) = \sum\nolimits_{n = 0}^\infty {{a_n}{\chi_n}(x)\quad \left( {{a_n} \ge 0,\sum\nolimits_{n = 0}^\infty {{a_n} = 1} } \right)} , where { c_n }_{n = 0}^￥ \left\{ {{\chi_n}} \right\}_{n = 0}^\infty is a system of multiplicative functions that are generalizations of the classical Walsh functions. For the semigroup M_P\mathcal{M}_{\mathcal{P}}, analogs of the well-known Khinchin theorems related to the arithmetic of a semigroup of probability measures in R ⁿ are true. We describe the class I₀(M_P)I_0(\mathcal{M}_{\mathcal{P}}) of functions without indivisible or nondegenerate idempotent divisors and construct a class of indecomposable functions that is dense in M_P\mathcal{M}_{\mathcal{P}} in the topology of uniform convergence.     

Large holes in sparse random graphs

We show that there is a function α(r) such that for each constantr≧3, almost everyr-regular graph onn vertices has a hole (vertex induced cycle) of size at least α(r)n asn→∞. We also show that there is a function β(c) such that forc>0 large enough,G _{n, p} ,p=c/n almost surely has a hole of size at least β(c)n asn→∞.     

Critical percolation on random regular graphs

The behavior of the random graph G(n,p) around the critical probability p_c = is well understood. When p = (1 + O(n^1/3))p_c the components are roughly of size n^2/3 and converge, when scaled by n^?2/3, to excursion lengths of a Brownian motion with parabolic drift. In particular, in this regime, they are not concentrated. When p = (1 ‐ ?(n))p_c with ?(n)n^1/3 →∞ (the subcritical regime) the largest component is concentrated around 2?^?2 log(?³n). When p = (1 + ?(n))p_c with ?(n)n^1/3 →∞ (the supercritical regime), the largest component is concentrated around 2?n and a duality principle holds: other component sizes are distributed as in the subcritical regime. Itai Benjamini asked whether the same phenomenon occurs in a random d‐regular graph. Some results in this direction were obtained by (Pittel, Ann probab 36 (2008) 1359–1389). In this work, we give a complete affirmative answer, showing that the same limiting behavior (with suitable d dependent factors in the non‐critical regimes) extends to random d‐regular graphs. © 2009 Wiley Periodicals, Inc. Random Struct. Alg., 2010     

Preconditioned HSS methods for the solution of non-Hermitian positive definite linear systems and applications to the discrete convection-diffusion equation

Summary. We study the role of preconditioning strategies recently developed for coercive problems in connection with a two-step iterative method, based on the Hermitian skew-Hermitian splitting (HSS) of the coefficient matrix, proposed by Bai, Golub and Ng for the solution of nonsymmetric linear systems whose real part is coercive. As a model problem we consider Finite Differences (FD) matrix sequences {A_n(a,p)}_n discretizing the elliptic (convection-diffusion) problem with being a plurirectangle of R^d with a(x) being a uniformly positive function and p(x) denoting the Reynolds function: here for plurirectangle we mean a connected union of rectangles in d dimensions with edges parallel to the axes. More precisely, in connection with preconditioned HSS/GMRES like methods, we consider the preconditioning sequence {P_n(a)}_n, P_n(a):= D_n^1/2(a)A_n(1,0) D_n^1/2(a) where D_n(a) is the suitably scaled main diagonal of A_n(a,0). If a(x) is positive and regular enough, then the preconditioned sequence shows a strong clustering at unity so that the sequence {P_n(a)}_n turns out to be a superlinear preconditioning sequence for {A_n(a,0)}_n where A_n(a,0) represents a good approximation of Re(A_n(a,p)) namely the real part of A_n(a,p). The computational interest is due to the fact that the preconditioned HSS method has a convergence behavior depending on the spectral properties of {P_n^-1(a)Re(A_n(a,p))}_n {P_n^-1(a)A_n(a,0)}_n: therefore the solution of a linear system with coefficient matrix A_n(a,p) is reduced to computations involving diagonals and to the use of fast Poisson solvers for {A_n(1,0)}_n.Some numerical experimentations confirm the optimality of the discussed proposal and its superiority with respect to existing techniques.Mathematics Subject Classification (1991): 65F10, 65N22, 15A18, 15A12, 47B65     

Hermite Interpolation and Sobolev Orthogonality

In this paper, we study orthogonal polynomials with respect to the bilinear form (f, g) _S = V(f) A V(g) ^T + <u, f ^(N) g ^(N)V(f) =(f(c ₀), f "(c ₀), ..., f ^{(n – 1)} ₀(c ₀), ..., f(c _p ), f "(c _p ), ..., f ^{(n – 1)} _p(c _p )) u is a regular linear functional on the linear space P of real polynomials, c ₀, c ₁, ..., c _p are distinct real numbers, n ₀, n ₁, ..., n _p are positive integer numbers, N=n ₀+n ₁+...+n _p , and A is a N × N real matrix with all its principal submatrices nonsingular. We establish relations with the theory of interpolation and approximation.     

On uniform estimates for solutions to quasi-linear differential equations

We obtain uniform estimates for positive solutions with the same domain to the equation

Strongly Singular Sturm – Liouville Problems

We consider a Sturm – Liouville operator Lu = —(r(t)u′)′ + p (t)u , where r is a (strictly) positive continuous function on ]a, b [ and p is locally integrable on ]a, b[. Let r₁(t) = (1/r) ds andchoose any c ∈]a, b[. We are interested in the eigenvalue problem Lu = λm(t)u, u (a) = u (b) = 0,and the corresponding maximal and anti .maximal principles, in the situation when 1/r ∈ L¹ (a, c),1 /r ∉ L¹ (c, b), pr₁ ∉ L¹ (a, c) and pr₁ ∉ L¹(c, b).     

On the Positive Definiteness of Certain Functions

For a coinmutative senugoup (S, +, *) with involution and a function f : S → [0, ∞), the set S(f) of those p ≥ 0 such that f^P is a positive definite function on S is a closed subsemigroup of [0, ∞) containing 0. For S = (IR, +, x* = -x) it may happen that S(f) = { kd : k ∈ N₀ } for some d > 0, and it may happen that S(f) = {0} ? [d, ∞) for some d > O. If α > 2 and if S = (?, +, n* = -n) and f(n) = e^?[n]α or S = (IN₀, +, n* = n) and f(n) = e^nα, then S(f) ∪ (0, c) = ? and [d, ∞) ? S(f) for some d ≥; c > 0. Although (with c maximal and d minimal) we have not been able to show c = d in all cases, this equality does hold if S = ? and α ≥ 3.4. In the last section we give sinipler proofs of previously known results concerning the positive definiteness of x → e^?||x||α on normed spaces.     

Simple two-sided anti-Lie-admissible algebras

For a Lie algebra L, a bilinear map is called a commutative cocycle if ψ(a, b) = ψ(b, a) and ψ([a, b], c) + ψ([b, a], c) + ψ([c, a], b) = 0 for any a, b, c ∈ L. We prove that any commutative cocycle of a simple Lie algebra of characteristic p ≠ 2, 3 is trivial if the rank of L is at least 2. In particular, any two-sided Alia algebra connected with a simple, finite-dimensional Lie algebra L is isomorphic to L, except for the case where L = sl ₂ . Translated from Sovremennaya Matematika i Ee Prilozheniya (Contemporary Mathematics and Its Applications), Vol. 60, Algebra, 2008.     

The asymptotic number of labeled connected graphs with a given number of vertices and edges

Let c(n, q) be the number of connected labeled graphs with n vertices and q ≤ N = (2n ) edges. Let x = q/n and k = q ? n. We determine functions w_k ? 1. a(x) and φ(x) such that c(n, q) ? w_k(q^N)eⁿφ(x)+a(x) uniformly for all n and q ≥ n. If ? > 0 is fixed, n→ ∞ and 4q > (1 + ?)n log n, this formula simplifies to c(n, q) ? (N^q) exp(–ne^?2q/n). on the other hand, if k = o(n^1/2), this formula simplifies to c(n, n + k) ? 1/2 w_k (3/π)^1/2 (e/12k)^k/2nⁿ?^(3k?1)/2.     

Convergence of ray sequences of Padé approximants for 2 f 1(a, 1; c; z), (c > a > 0)

The Padé table of ₂ F ₁(a, 1; c; z) is normal for c > a > 0 (cf. [4]). For m ≥ n - 1 and c ? Z^-, the denominator polynomial Q _mn (z) in the [m/n] Padé approximant P _mn (z)/Q _mn (z) for ₂ F ₁(a, 1; c; z) and the remainder term Q _mn (z)₂ F ₁(a, 1; c; z)-P_mn (z) were explicitly evaluated by Padé (cf. [2], [6] or [9]). We show that for c > a > 0 and m ≥ n - 1, the poles of P_mn (z)/Q_mn (z) lie on the cut (1,∞). We deduce that the sequence of approximants P_mn (z)/Q_mn (z) converges to ₂ F ₁(a, 1; c; z) as m → ∞, n/m → ρ with 0 < ρ ≤ 1, uniformly on compact subsets of the unit disc |z| < 1 for c > a > 0.     

Extreme Points and Minimal Outer Area Problem for Meromorphic Univalent Functions

For 0 < p < 1, letS_pdenote the class of functionsf(z) meromorphic univalent in the unit disk with the normalizationf(0) = 0,f′(0) = 1, andf(p) = ∞. LetS_p(a) be the subclass ofS_pwith the fixed residuea. In this note we determine the extreme points of the classS_p(a). As an application, we solve the problem of minimizing the outer area overS_p(a), which was posed by S. Zemyan (J. Analyse Math.39, 1981, 11–23).     

On the set of limit points of the partial sums of series rearranged by a given divergent permutation

We give a new characterization of divergent permutations. We prove that for any divergent permutation p, any closed interval I of (the 2-point compactification of ) and any real number sI, there exists a series ∑a_n of real terms convergent to s such that I=σa_p(n) (where σa_p(n) denotes the set of limit points of the partial sums of the series ∑a_p(n)). We determine permutations p of for which there exists a conditionally convergent series ∑a_n such that ∑a_p(n)=+∞. If the permutation p of possesses the last property then we prove that for any and there exists a series ∑a_n convergent to α and such that σa_p(n)=[β,+∞]. We show that for any countable family P of divergent permutations there exist conditionally convergent series ∑a_n and ∑b_n such that any series of the form ∑a_p(n) with pP is convergent to the sum of ∑a_n, while for every pP.     

Almost universal graphs

We study the question as to when a random graph with n vertices and m edges contains a copy of almost all graphs with n vertices and cn/2 edges, c constant. We identify a “phase transition” at c = 1. For c < 1, m must grow slightly faster than n, and we prove that m = O(n log log n/log log log n) is sufficient. When c > 1, m must grow at a rate m = n^1+a, where a = a(c) > 0 for every c > 1, and a(c) is between and for large enough c. © 2006 Wiley Periodicals, Inc. Random Struct. Alg., 2006