A greedy approximation algorithm for the group Steiner problem

In the group Steiner problem we are given an edge-weighted graph G=(V,E,w) and m subsets of vertices . Each subset g_i is called a group and the vertices in ?_ig_i are called terminals. It is required to find a minimum weight tree that contains at least one terminal from every group.We present a poly-logarithmic ratio approximation for this problem when the input graph is a tree. Our algorithm is a recursive greedy algorithm adapted from the greedy algorithm for the directed Steiner tree problem [Approximating the weight of shallow Steiner trees, Discrete Appl. Math. 93 (1999) 265-285, Approximation algorithms for directed Steiner problems, J. Algorithms 33 (1999) 73-91]. This is in contrast to earlier algorithms that are based on rounding a linear programming based relaxation for the problem [A polylogarithmic approximation algorithm for the Group Steiner tree problem, J. Algorithms 37 (2000) 66-84, preliminary version in Proceedings of SODA, 1998 pp. 253-259, On directed Steiner trees, Proceedings of SODA, 2002, pp. 59-63]. We answer in positive a question posed in [A polylogarithmic approximation algorithm for the Group Steiner tree problem, J. Algorithms 37 (2000) 66-84, preliminary version in Proceedings of SODA, 1998 pp. 253-259] on whether there exist good approximation algorithms for the group Steiner problem that are not based on rounding linear programs. For every fixed constant ε>0, our algorithm gives an approximation in polynomial time. Approximation algorithms for trees can be extended to arbitrary undirected graphs by probabilistically approximating the graph by a tree. This results in an additional multiplicative factor of in the approximation ratio, where |V| is the number of vertices in the graph. The approximation ratio of our algorithm on trees is slightly worse than the ratio of O(log(max_i|g_i|)·logm) provided by the LP based approaches.     

An improved approximation algorithm for requirement cut

This note presents improved approximation guarantees for the requirement cut problem: given an n-vertex edge-weighted graph G=(V,E), and g groups of vertices X₁,…,X_g⊆V with each group X_i having a requirement r_i between 0 and |X_i|, the goal is to find a minimum cost set of edges whose removal separates each group X_i into at least r_i disconnected components. We give a tight Θ(logg) approximation ratio for this problem when the underlying graph is a tree, and show how this implies an O(logk⋅logg) approximation ratio for general graphs, where .     

An algorithmic proof of Tutte's f-factor theorem

Let G be a multigraph on n vertices, possibly with loops. An f-factor is a subgraph of G with degree f_i at the ith vertex for i = 1, 2,…, n. Tutte's f-factor theorem is proved by providing an algorithm that either finds an f-factor or shows that it does not exist and does this in O(n³) operations. Note that the complexity bound is independent of the number of edges of G and the degrees f_i. The algorithm is easily altered to handle the problem of looking for a symmetric integral matrix with given row and column sums by assigning loops degree one. A (g,f)-factor is a subgraph of G with degree d_i at the ith vertex, where g_i ? d_i ? f_i, for i = 1,2,…, n. Lovasz's (g,f)-factor theorem is proved by providing an O(n³) algorithm to either find a (g,f)-factor or show one does not exist.     

A survey of published programs for best approximation

We list and discuss published programs for best approximation of functions by linear and nonlinear families in all standard forms.In this note we list and discuss the published programs for obtaining best approximations. Let X be a set on which we wish to approximate. Most sets will be finite (an equivalent terms is discrete). Let 6 6 be a norm on the continuous functions on X. Let G be a familiy of continuous functions on X. For a given basis {φ₁…,φ_n}, the linear family G is the set of all functions of the form

$\text{L(A,x)=}\text{∑}\text{k=1}\text{n}\text{a}{}_{\mathrm{k}}\text{φ}{}_{\mathrm{k}}\text{(x)}$

The problem of best approximation is given a continuous function f, to find g¹ to minimize e(g) = ∥f-g∥ over g?G. Such g¹ is called a best approximation to f.Discrete linear approximation problems are sometimes formulated as solution of an overdetermined system of linear equations Ax = b with respect to a norm where a_ij = φ_j(x_i) and b_i = f(x)_i).     

What Is the Complexity of Solution-Restricted Operator Equations?

We study the worst case complexity of operator equations Lu = f where L: G → X is a bounded linear injection of normed linear spaces. Past work on the complexity of such problems has generally required the class F of problem elements f to be the unit ball of X. However, there are many problems for which this choice of F yields unsatisfactory results. Mixed elliptic—hyperbolic problems are one example. the difficulty being that our technical tools are nor strong enough to give good complexity bounds. Ill-posed problems are another example. because we know that the complexity of computing finite-error approximations is infinite if F is a ball in X. In this paper, we pursue another idea. Rather than directly restrict the class F of problem elements f, we will consider problems that are solution-restricted: i.e., we restrict the class U of solution elements u. In particular, we assume that U is the unit hall of a normed linear space W that is densely, continuously embedded in G. The main idea is that our problem can now be reduced to the standard approximation problem of approximating the embedding of W into G.This allows us to characterize optimal information and algorithms for our problem..We use this idea to study three problems: the Tricomi problem (a mixed hyperbolic— elliptic problem arising in the study of transonic flow), the inverse finite Laplace transform (an ill-posed problem arising. e.g.. in geomathematics), and the backwards heat equation. We determine the problem complexity and derive nearly optimal algorithms for each of these problems.     

Approximation of functions of finite variation by superpositions of a Sigmoidal function

The aim of this note is to generalize a result of Barron [1] concerning the approximation of functions, which can be expressed in terms of the Fourier transform, by superpositions of a fixed sigmoidal function. In particular, we consider functions of the type h(x) = ∫ℝ^d ƒ (〈t, x〉)dμ(t), where μ is a finite Radon measure on ℝ^d and ƒ : ℝ → ℂ is a continuous function with bounded variation in ℝ We show (Theorem 2.6) that these functions can be approximated in L₂-norm by elements of the set G_n = {Σ_i=0^staggeredn c_ig(〈a_i, x〉 + b_i) : a_iℝ^d, b_i, c_iℝ}, where g is a fixed sigmoidal function, with the error estimated by C/n^1/2, where C is a positive constant depending only on f. The same result holds true (Theorem 2.9) for f : ℝ → ℂ satisfying the Lipschitz condition under an additional assumption that ∫ℝ^d6t6e^d|u(t)| > ∞     

Free generating sets of lattice-ordered abelian groups

We prove that the generators g₁,…,g_n of a lattice-ordered abelian group G form a free generating set iff each ?-ideal generated by any n−1 linear combinations of the g_i is strictly contained in some maximal ?-ideal of G.     

An addition theorem for finite Abelian groups

If g₁, g₂, …, g_2n?1 is a sequence of 2n ? 1 elements in an Abelian group G of order n, it is known that there are n distinct indices i₁, i₂, …, i_n such that 0 = g_i1 + g_i2 + ? + g_in. In this paper a suitably general condition on the sequence is given which insures that every element g in G has a representation g = g_i1 + g_i2 + ? + g_in as the sum of n terms of the sequence.     

Recursive linear orders with recursive successivities

A successivity in a linear order is a pair of elements with no other elements between them. A recursive linear order with recursive successivities $\text{U}$ is recursively categorical if every recursive linear order with recursive successivities isomorphic to $\text{U}$ is in fact recursively isomorphic to $\text{U}$. We characterize those recursive linear orders with recursive successivities that are recursively categorical as precisely those with order type k₁+g₁+k₂+g₂+…+g_n-1+k_n where each k_n is a finite order type, non-empty for i?{2,…,n-1} and each g_i is an order type from among {ω,ω^*,ω+ω^*}∪{k·η:k<ω}.     

Metrics and undirected cuts

Suppose thatG is an undirected graph whose edges have nonnegative integer-valued lengthsl(e), and that {s ₁,t ₁},?, {s _m ,t _m } are pairs of its vertices. Can one assign nonnegative weights to the cuts ofG such that, for each edgee, the total weight of cuts containinge does not exceedl(e) and, for eachi, the total weight of cuts ‘separating’s _i andt _i is equal to the distance (with respect tol) betweens _i andt _i ? Using linear programming duality, it follows from Papernov's multicommodity flow theorem that the answer is affirmative if the graph induced by the pairs {s ₁,t ₁},?, {s _m ,t _m } is one of the following: (i) the complete graph with four vertices, (ii) the circuit with five vertices, (iii) a union of two stars. We prove that if, in addition, each circuit inG has an even length (with respect tol) then there exists a suitable weighting of the cuts with the weights integer-valued; moreover, an algorithm of complexity O(n ³) (n is the number of vertices ofG) is developed for solving such a problem. Also a class of metrics decomposable into a nonnegative linear combination of cut-metrics is described, and it is shown that the separation problem for cut cones isNP-hard.     

Complexity of Neural Network Approximation with Limited Information: A Worst Case Approach

In neural network theory the complexity of constructing networks to approximate input-output functions is of interest. We study this in the more general context of approximating elements f of a normed space F using partial information about f. We assume information about f and the size of the network are limited, as is typical in radial basis function networks. We show complexity can be essentially split into two independent parts, information ε-complexity and neural ε-complexity. We use a worst case setting, and integrate elements of information-based complexity and nonlinear approximation. We consider deterministic and/or randomized approximations using information possibly corrupted by noise. The results are illustrated by examples including approximation by piecewise polynomial neural networks.     

The Erd?s-Ginzburg-Ziv theorem for finite solvable groups

Let G be a non-cyclic finite solvable group of order n, and let S=(g₁,…,g_k) be a sequence of k elements (repetition allowed) in G. In this paper we prove that if , then there exist some distinct indices i₁,i₂,…,i_n such that the product g_i1g_i2?g_in=1. This result substantially improves the Erd?s-Ginzburg-Ziv theorem and other existing results.     

Some Complexity Results for Polynomial Ideals

In this paper, we survey some of our new results on the complexity of a number of problems related to polynomial ideals. We consider multivariate polynomials over some ring, like the integers or the rationals. For instance, a polynomial ideal membership problem is a (w+ 1)-tupleP= (f,g₁,g₂, …g_wwherefand theg_iare multivariate polynomials, and the problem is to determine whetherfis in the ideal generated by theg_i. For polynomials over the integers or rationals, this problem is known to be exponential space complete. We discuss further complexity results for problems related to polynomial ideals, like the word and subword problems for commutative semigroups, a quantitative version of Hilbert's Nullstellensatz in a complexity theoretic version, and problems concerning the computation of reduced polynomials and Gröbner bases.     

Weighted rational approximation in the complex plane

Given a triple (G, W, γ) of an open bounded set G in the complex plane, a weight function W(z) which is analytic and different from zero in G, and a number γ with 0 ≤ γ ≤ 1, we consider the problem of locally uniform rational approximation of any function ƒ(z), which is analytic in G, by weighted rational functions W^m_i+n_i(z)R_mi, n_i(z)_{i = 0}^∞, where R_mi, n_i(z) = P_mi(z)/Q_ni(z) with deg P_mi ≤ m_i and deg Q_ni ≤ n_i for all i ≥ 0 and where m_i + n_i → ∞ as i → ∞ such that lim m_i/[m_i + n_i] = γ. Our main result is a necessary and sufficient condition for such an approximation to be valid. Applications of the result to various classical weights are also included.     

Examples of cell-like decompositions of the infinite-dimensional manifolds σ and Σ

Denote by σ the subspace of Hilbert space {(x_i)?l₂:x_i=0 for all but finitely many i}. Examples of cell-like decompositions of σ are constructed that have decomposition spaces that are not homeomorphic to σ. At one extreme is a cell-like decomposition G of σ produced using ghastly finite dimensional examples such that the decomposition space σ?G contains no embedded 2-cell but (σ?G)×$\text{R}$ is homeomorphic to σ. At the other extreme is a cell-like decomposition G of σ satisfying: (a) the nondegeneracy set N_G={g?G:g≠point} consists of countably many arcs (necessarily tame); (b) the nondegeneracy set N_G is a closed subset of the decomposition space σ?G; (c) each map f:B²→σ?G of a 2-cell into σ?G can be approximated arbitrarily closely by an embedding; (d) σ?G is not homeomorphic to σ but (σ?G)×$\text{R}$ is homeomorphic to σ. The fact that both conditions (a) and (b) can be satisfied (and have (d) hold) is directly attributable to σ’s incompleteness as a topological space.     

A description of Baer-Suzuki type of the solvable radical of a finite group

We obtain the following characterization of the solvable radical R(G) of any finite group G: R(G) coincides with the collection of all g∈G such that for any 3 elements a₁,a₂,a₃∈G the subgroup generated by the elements , i=1,2,3, is solvable. In particular, this means that a finite group G is solvable if and only if in each conjugacy class of G every 4 elements generate a solvable subgroup. The latter result also follows from a theorem of P. Flavell on {2,3}^′-elements in the solvable radical of a finite group (which does not use the classification of finite simple groups).     

Greedy Approximation of High-Dimensional Ornstein–Uhlenbeck Operators

We investigate the convergence of a nonlinear approximation method introduced by Ammar et?al. (J. Non-Newtonian Fluid Mech. 139:153–176, 2006) for the numerical solution of high-dimensional Fokker–Planck equations featuring in Navier–Stokes–Fokker–Planck systems that arise in kinetic models of dilute polymers. In the case of Poisson’s equation on a rectangular domain in ?², subject to a homogeneous Dirichlet boundary condition, the mathematical analysis of the algorithm was carried out recently by Le Bris, Lelièvre and Maday (Const. Approx. 30:621–651, 2009), by exploiting its connection to greedy algorithms from nonlinear approximation theory, explored, for example, by DeVore and Temlyakov (Adv. Comput. Math. 5:173–187, 1996); hence, the variational version of the algorithm, based on the minimization of a sequence of Dirichlet energies, was shown to converge. Here, we extend the convergence analysis of the pure greedy and orthogonal greedy algorithms considered by Le Bris et al. to a technically more complicated situation, where the Laplace operator is replaced by an Ornstein–Uhlenbeck operator of the kind that appears in Fokker–Planck equations that arise in bead–spring chain type kinetic polymer models with finitely extensible nonlinear elastic potentials, posed on a high-dimensional Cartesian product configuration space D=D ₁×?×D _N contained in ? ^Nd , where each set D _i , i=1,…,N, is a bounded open ball in ? ^d , d=2,3.