Vol 94, No 3 (2016)

The problem of nonparametric estimation of a signal function by thresholding the coefficients of its wavelet decomposition is considered. In models with various noise distributions, asymptotically optimal thresholds and orders of the loss functions are calculated on the basis of probabilities of errors in the calculation of wavelet coefficients.

The local clustering coefficients of preferential attachment models are analyzed. Previously, a general approach to preferential attachment was proposed (the PA-class was introduced); it was shown that the degree distribution in all models of the PA-class obeys a power law. The global clustering coefficient was also analyzed, and a lower bound for the mean local clustering coefficient was found. In the paper, new results are obtained by analyzing the local clustering coefficients of models of the PA-class. Namely, the behavior of the mean value C₂(n, d) of local clustering over vertices of degree d is studied.

In the present paper, we consider word maps w: G^m → G and word maps with constants w_Σ: G^m → G of a simple algebraic group G, where w is a nontrivial word in the free group F_m of rank m, w_Σ = w₁σ₁w₂ ··· w_rσ_rw_{r + 1}, w₁, …, w_{r + 1} ∈ F_m, w₂, …, w_r ≠ 1, Σ = {σ₁, …, σ_r | σ_i ∈ GZ(G)}. We present results on the images of such maps, in particular, we prove a theorem on the dominance of “general” word maps with constants, which can be viewed as an analogue of a well-known theorem of Borel on the dominance of genuine word maps. Besides, we establish a relationship between the existence of unipotents in the image of a word map and the structure of the representation variety R(Γ_w, G) of the group Γ_w = F_m/<w>.

Volterra integrodifferential equations with unbounded operator coefficients in a Hilbert space that are operator models of integrodifferential equations arising in viscoelasticity theory are studied. These equations are shown to be well-posed in Sobolev spaces of vector functions, and spectral analysis is applied to the operator functions that are the symbols of the given equations.

The oblique derivative problem for the heat equation is considered in a model formulation with a boundary function that can be discontinuous and with the boundary condition understood as the limit in the normal direction almost everywhere on the lateral boundary of the domain. An example is given showing that the solution is not unique in this formulation. A solution is sought in the parabolic Zygmund space H₁, which is an analogue of the parabolic Hölder space for an integer smoothness exponent. A subspace of H₁ is introduced in which the existence and uniqueness of the solution is proved under suitable assumptions about the data of the problem.

Representations of Schrödinger semigroups and groups by Feynman iterations are studied. The compactness, rather than convergence, of the sequence of Feynman iterations is considered. Approximations of solutions of the Cauchy problem for the Schrödinger equation by Feynman iterations are investigated. The Cauchy problem for the Schrödinger equation under consideration is ill-posed. From the point of view of the approach of the paper, this means that the problem has no solution in the sense of integral identity for some initial data. The well-posedness of the Cauchy problem can be recovered by extending the operator to a selfadjoint one; however, there exists continuum many such extensions. Feynman iterations whose partial limits are the solutions of all Cauchy problems obtained for various self-adjoint extensions are studied.

The polynomial sub-Riemannian differentiability of classes of mappings of Carnot groups and graphs is proved. Examples of polynomial sub-Riemannian differentials preserving Hausdorff dimension are given.

The solvability (in classical sense) of the Bitsadze–Samarskii nonlocal initial–boundary value problem for a one-dimensional (in x) second-order parabolic system in a semibounded domain with a nonsmooth lateral boundary is proved by applying the method of boundary integral equations. The only condition imposed on the right-hand side of the nonlocal boundary condition is that it has a continuous derivative of order 1/2 vanishing at t = 0. The smoothness of the solution is studied.

A transformation of a discrete-time martingale with conditionally Gaussian increments into a sequence of i.i.d. standard Gaussian random variables is proposed as based on a sequence of stopping times constructed using the quadratic variation. It is shown that sequential estimators for the parameters in AR(1) and generalized first-order autoregressive models have a nonasymptotic normal distribution.

An embedding of the Sobolev spaces W_p^s (ℝⁿ) in Lizorkin-type spaces of locally integrable functions of smoothness zero is obtained; a similar assertion for Riesz and Bessel potentials is presented. The embedding theorem is extended to Sobolev spaces on irregular domains in n-dimensional Euclidean space. The statement of the theorem depends on geometric parameters of the domain of functions.

A relationship between the continued fraction expansion of the quadratic irrationalities of hyperelliptic fields and the Mumford polynomials determining addition in the group of divisor classes on a hyperelliptic curve is described. A theorem on the equivalence of the quasi-periodicity of a quadratic irrationality and the existence of a point of finite order is proved; results on the symmetry of the quasi-period and estimates of its length are obtained.

Solutions to a partial differential equation of the third order containing the modular nonlinearity are studied. The model describes, in particular, elastic waves in media with weak high-frequency dispersion and with different response to tensile and compressive stresses. This equation is linear for solutions preserving their sign. Nonlinear phenomena only manifest themselves to alternating solutions. Stationary solutions in the form of solitary waves or solitons are found. It is shown how the linear periodic wave becomes nonlinear after exceeding a certain critical value of the amplitude, and how it transforms into a soliton with further increase in the amplitude.

Some inadequate results may appear in the DEA models as in any other mathematical model. In the DEA scientific literature several methods were proposed to deal with these difficulties. In our previous paper, we introduced the notion of terminal units. It was also substantiated that only terminal units form necessary and sufficient sets of units for smoothing the frontier. Moreover, some relationships were established between terminal units and other sets of units that were proposed for improving the frontier. In this paper we develop a general algorithm for improving the frontier. The construction of algorithm is based on the notion of terminal units. Our theoretical results are verified by computational results using real-life data sets and also confirmed by graphical examples.