All issues

In the current paper, we study a Galerkin–Petrov method for a parabolic equations of higher order in domain with a moving boundary. Asymptotic estimates for the convergence rate of approximate solutions are obtained.

The generalization of Margolus’s block cellular automaton on a hexagonal grid is formulated. Statistical analysis of the results of probabilistic cellular automation for vast variety of this scheme solving the test task of diffusion is done. It is shown that the choice of the hexagon blocks is 25% more efficient than Y-blocks. It is shown that the algorithms have polynomial complexity, and the polynom degree lies within 0.6÷0.8 for parallel computer, and in the range 1.5÷1.7 for serial computer. The effects of embedded into automaton’s field defective cells on the rate of convergence are studied also.

The second part of paper is devoted to final study of three classic partial differential equations (Laplace, Diffusion and Wave) solution using simple numerical methods in terms of Cellular Automata. Specificity of this solution has been shown by different examples, which are related to the hexagonal grid. Also the next statements that are mentioned in the first part have been proved: the matter conservation law and the offensive effect of excessive hexagonal symmetry.

From the point of CA view diffusion equation is the most important. While solving of diffusion equation at the infinite time interval we can find solution of boundary value problem of Laplace equation and if we introduce vector-variable we will solve wave equation (at least, for scalar). The critical requirement for the sampling of the boundary conditions for CA-cells has been shown during the solving of problem of circular membrane vibrations with Neumann boundary conditions. CA-calculations using the simple scheme and Margolus rotary-block mechanism were compared for the quasione-dimensional problem “diffusion in the half-space”. During the solving of mixed task of circular membrane vibration with the fixed ends in a classical case it has been shown that the simultaneous application of the Crank–Nicholson method and taking into account of the second-order terms is allowed to avoid the effect of excessive hexagonal symmetry that was studied for a simple scheme.

By the example of the centrally symmetric Neumann problem a new method of spatial derivatives introducing into the postfix CA procedure, which is reflecting the time derivatives (on the base of the continuity equation) was demonstrated. The value of the constant that is related to these derivatives has been empirically found in the case of central symmetry. The low rate of convergence and accuracy that limited within the boundaries of the sample, in contrary to the formal precision of the method (4-th order), prevents the using of the CAmethods for such problems. We recommend using multigrid method. During the solving of the quasi-diffusion equations (two-dimensional CA) it was showing that the rotary-block mechanism of CA (Margolus mechanism) is more effective than simple CA.

In the article a theory of the implicit iterative line-by-line recurrence method for solving the systems of finite-difference equations which arise as a result of approximation of the two-dimensional elliptic differential equations on a regular grid is stated. On the one hand, the high effectiveness of the method has confirmed in practice. Some complex test problems, as well as several problems of fluid flow and heat transfer of a viscous incompressible liquid, have solved with its use. On the other hand, the theoretical provisions that explain the high convergence rate of the method and its stability are not yet presented in the literature. This fact is the reason for the present investigation. In the paper, the procedure of equivalent and approximate transformations of the initial system of linear algebraic equations (SLAE) is described in detail. The transformations are presented in a matrix-vector form, as well as in the form of the computational formulas of the method. The key points of the transformations are illustrated by schemes of changing of the difference stencils that correspond to the transformed equations. The canonical form of the method is the goal of the transformation procedure. The correctness of the method follows from the canonical form in the case of the solution convergence. The estimation of norms of the matrix operators is carried out on the basis of analysis of structures and element sets of the corresponding matrices. As a result, the convergence of the method is proved for arbitrary initial vectors of the solution of the problem.

The norm of the transition matrix operator is estimated in the special case of weak restrictions on a desired solution. It is shown, that the value of this norm decreases proportionally to the second power (or third degree, it depends on the version of the method) of the grid step of the problem solution area in the case of transition matrix order increases. The necessary condition of the method stability is obtained by means of simple estimates of the vector of an approximate solution. Also, the estimate in order of magnitude of the optimum iterative compensation parameter is given. Theoretical conclusions are illustrated by using the solutions of the test problems. It is shown, that the number of the iterations required to achieve a given accuracy of the solution decreases if a grid size of the solution area increases. It is also demonstrated that if the weak restrictions on solution are violated in the choice of the initial approximation of the solution, then the rate of convergence of the method decreases essentially in full accordance with the deduced theoretical results.

The work belongs to the direction of experimental mathematics, which investigates the properties of mathematical objects by the computing facilities of a computer. The base is an exponential map, its topological properties (Cantor's bouquets) differ from properties of polynomial and rational complex-valued functions. The subject of the study are the character and features of the Fatou and Julia sets, as well as the equilibrium points and orbits of the zero of three iterated complex-valued mappings: $f:z \to (1+ \mu) \exp (iz)$, $g : z \to \big(1+ \mu |z - z^*|\big) \exp (iz)$, $h : z \to \big(1+ \mu (z - z^* )\big) \exp (iz)$, with $z,\mu \in \mathbb{C}$, $z^* : \exp (iz^*) = z^*$. For a quasilinear map g having no analyticity characteristic, two bifurcation transitions were discovered: the creation of a new equilibrium point (for which the critical value of the linear parameter was found and the bifurcation consists of “fork” type and “saddle”-node transition) and the transition to the radical transformation of the Fatou set. A nontrivial character of convergence to a fixed point is revealed, which is associated with the appearance of “valleys” on the graph of convergence rates. For two other maps, the monoperiodicity of regimes is significant, the phenomenon of “period doubling” is noted (in one case along the path $39\to 3$, in the other along the path $17\to 2$), and the coincidence of the period multiplicity and the number of sleeves of the Julia spiral in a neighborhood of a fixed point is found. A rich illustrative material, numerical results of experiments and summary tables reflecting the parametric dependence of maps are given. Some questions are formulated in the paper for further research using traditional mathematics methods.

In this work we consider Monteiro – Svaiter accelerated hybrid proximal extragradient (A-HPE) framework and accelerated Newton proximal extragradient (A-NPE) framework. The last framework contains an optimal method for rather smooth convex optimization problems with second-order oracle. We generalize A-NPE framework for higher order derivative oracle (schemes). We replace Newton’s type step in A-NPE that was used for auxiliary problem by Newton’s regularized (tensor) type step (Yu. Nesterov, 2018). Moreover we generalize large step A-HPE/A-NPE framework by replacing Monteiro – Svaiter’s large step condition so that this framework could work for high-order schemes. The main contribution of the paper is as follows: we propose optimal highorder methods for convex optimization problems. As far as we know for that moment there exist only zero, first and second order optimal methods that work according to the lower bounds. For higher order schemes there exists a gap between the lower bounds (Arjevani, Shamir, Shiff, 2017) and existing high-order (tensor) methods (Nesterov – Polyak, 2006; Yu.Nesterov, 2008; M. Baes, 2009; Yu.Nesterov, 2018). Asymptotically the ratio of the rates of convergences for the best existing methods and lower bounds is about 1.5. In this work we eliminate this gap and show that lower bounds are tight. We also consider rather smooth strongly convex optimization problems and show how to generalize the proposed methods to this case. The basic idea is to use restart technique until iteration sequence reach the region of quadratic convergence of Newton method and then use Newton method. One can show that the considered method converges with optimal rates up to a logarithmic factor. Note, that proposed in this work technique can be generalized in the case when we can’t solve auxiliary problem exactly, moreover we can’t even calculate the derivatives of the functional exactly. Moreover, the proposed technique can be generalized to the composite optimization problems and in particular to the constraint convex optimization problems. We also formulate a list of open questions that arise around the main result of this paper (optimal universal method of high order e.t.c.).

In this work we consider a possibility to use the conception of $(\delta, L)$-model of a function for optimization tasks, whereby solving a primal problem there is a necessity to recover a solution of a dual problem. The conception of $(\delta, L)$-model is based on the conception of $(\delta, L)$-oracle which was proposed by Devolder–Glineur–Nesterov, herewith the authors proposed approximate a function with an upper bound using a convex quadratic function with some additive noise $\delta$. They managed to get convex quadratic upper bounds with noise even for nonsmooth functions. The conception of $(\delta, L)$-model continues this idea by using instead of a convex quadratic function a more complex convex function in an upper bound. Possibility to recover the solution of a dual problem gives great benefits in different problems, for instance, in some cases, it is faster to find a solution in a primal problem than in a dual problem. Note that primal-dual methods are well studied, but usually each class of optimization problems has its own primal-dual method. Our goal is to develop a method which can find solutions in different classes of optimization problems. This is realized through the use of the conception of $(\delta, L)$-model and adaptive structure of our methods. Thereby, we developed primal-dual adaptive gradient method and fast gradient method with $(\delta, L)$-model and proved convergence rates of the methods, moreover, for some classes of optimization problems the rates are optimal. The main idea is the following: we find a dual solution to an approximation of a primal problem using the conception of $(\delta, L)$-model. It is much easier to find a solution to an approximated problem, however, we have to do it in each step of our method, thereby the principle of “divide and conquer” is realized.

In this paper, we consider conditional gradient methods for optimizing strongly convex functions. These are methods that use a linear minimization oracle, which, for a given vector $p \in \mathbb{R}^n$, computes the solution of the subproblem

$$\text{Argmin}_{x\in X}{\langle p,\,x \rangle}.$$ There are a variety of conditional gradient methods that have a linear convergence rate in a strongly convex case. However, in all these methods, the dimension of the problem is included in the rate of convergence, which in modern applications can be very large. In this paper, we prove that in the strongly convex case, the convergence rate of the conditional gradient methods in the best case depends on the dimension of the problem $n$ as $\widetilde {\Omega} \left(\!\sqrt {n}\right)$. Thus, the conditional gradient methods may turn out to be ineffective for solving strongly convex optimization problems of large dimensions.

Also, the application of conditional gradient methods to minimization problems of a quadratic form is considered. The effectiveness of the Frank – Wolfe method for solving the quadratic optimization problem in the convex case on a simplex (PageRank) has already been proved. This work shows that the use of conditional gradient methods to solve the minimization problem of a quadratic form in a strongly convex case is ineffective due to the presence of dimension in the convergence rate of these methods. Therefore, the Shrinking Conditional Gradient method is considered. Its difference from the conditional gradient methods is that it uses a modified linear minimization oracle. It's an oracle, which, for a given vector $p \in \mathbb{R}^n$, computes the solution of the subproblem $$\text{Argmin}\{\langle p, \,x \rangle\colon x\in X, \;\|x-x_0^{}\| \leqslant R \}.$$ The convergence rate of such an algorithm does not depend on dimension. Using the Shrinking Conditional Gradient method the complexity (the total number of arithmetic operations) of solving the minimization problem of quadratic form on a $\infty$-ball is obtained. The resulting evaluation of the method is comparable to the complexity of the gradient method.

Recently, to describe various mathematical models of physical processes, fractional differential calculus has been widely used. In this regard, much attention is paid to partial differential equations of fractional order, which are a generalization of partial differential equations of integer order. In this case, various settings are possible.

Loaded differential equations in the literature are called equations containing values of a solution or its derivatives on manifolds of lower dimension than the dimension of the definitional domain of the desired function. Currently, numerical methods for solving loaded partial differential equations of integer and fractional orders are widely used, since analytical solving methods for solving are impossible. A fairly effective method for solving this kind of problem is the finite difference method, or the grid method.

We studied the initial-boundary value problem in the rectangle $\overline{D}=\{(x,\,t)\colon 0\leqslant x\leqslant l,\;0\leqslant t\leqslant T\}$ for the loaded differential heat equation with composition fractional derivative of Riemann – Liouville and Caputo – Gerasimov and with boundary conditions of the first and third kind. We have gotten an a priori assessment in differential and difference interpretations. The obtained inequalities mean the uniqueness of the solution and the continuous dependence of the solution on the input data of the problem. A difference analogue of the composition fractional derivative of Riemann – Liouville and Caputo –Gerasimov order $(2-\beta )$ is obtained and a difference scheme is constructed that approximates the original problem with the order $O\left(\tau +h^{2-\beta } \right)$. The convergence of the approximate solution to the exact one is proven at a rate equal to the order of approximation of the difference scheme.