All issues

Conjugate gradient algorithms represent an important class of unconstrained optimization algorithms with strong local and global convergence properties and simple memory requirements. These algorithms have advantages that place them between the steep regression method and Newton’s algorithm because they require calculating the first derivatives only and do not require calculating and storing the second derivatives that Newton’s algorithm needs. They are also faster than the steep descent algorithm, meaning that they have overcome the slow convergence of this algorithm, and it does not need to calculate the Hessian matrix or any of its approximations, so it is widely used in optimization applications. This study proposes a novel method for image restoration by fusing the convex combination method with the hybrid (CG) method to create a hybrid three-term (CG) algorithm. Combining the features of both the Fletcher and Revees (FR) conjugate parameter and the hybrid Fletcher and Revees (FR), we get the search direction conjugate parameter. The search direction is the result of concatenating the gradient direction, the previous search direction, and the gradient from the previous iteration. We have shown that the new algorithm possesses the properties of global convergence and descent when using an inexact search line, relying on the standard Wolfe conditions, and using some assumptions. To guarantee the effectiveness of the suggested algorithm and processing image restoration problems. The numerical results of the new algorithm show high efficiency and accuracy in image restoration and speed of convergence when used in image restoration problems compared to Fletcher and Revees (FR) and three-term Fletcher and Revees (TTFR).

Mathematical models of many natural processes are described by partial differential equations with singular solutions. Classical numerical methods for determination of approximate solution to such problems are inefficient. In the present paper a boundary value problem for vector wave equation in L-shaped domain is considered. The presence of reentrant corner of size $3\pi/2$ on the boundary of computational domain leads to the strong singularity of the solution, i.e. it does not belong to the Sobolev space $H^1$ so classical and special numerical methods have a convergence rate less than $O(h)$. Therefore in the present paper a special weighted set of vector-functions is introduced. In this set the solution of considered boundary value problem is defined as $R_ν$-generalized one.

For numerical determination of the $R_ν$-generalized solution a weighted vector finite element method is constructed. The basic difference of this method is that the basis functions contain as a factor a special weight function in a degree depending on the properties of the solution of initial problem. This allows to significantly raise a convergence speed of approximate solution to the exact one when the mesh is refined. Moreover, introduced basis functions are solenoidal, therefore the solenoidal condition for the solution is taken into account precisely, so the spurious numerical solutions are prevented.

Results of numerical experiments are presented for series of different type model problems: some of them have a solution containing only singular component and some of them have a solution containing a singular and regular components. Results of numerical experiment showed that when a finite element mesh is refined a convergence rate of the constructed weighted vector finite element method is $O(h)$, that is more than one and a half times better in comparison with special methods developed for described problem, namely singular complement method and regularization method. Another features of constructed method are algorithmic simplicity and naturalness of the solution determination that is beneficial for numerical computations.

The article considers minimization of the expectation of convex function. Problems of this type often arise in machine learning and a variety of other applications. In practice, stochastic gradient descent (SGD) and similar procedures are usually used to solve such problems. We propose to use the ellipsoid method with mini-batching, which converges linearly and can be more efficient than SGD for a class of problems. This is verified by our experiments, which are publicly available. The algorithm does not require neither smoothness nor strong convexity of the objective to achieve linear convergence. Thus, its complexity does not depend on the conditional number of the problem. We prove that the method arrives at an approximate solution with given probability when using mini-batches of size proportional to the desired accuracy to the power −2. This enables efficient parallel execution of the algorithm, whereas possibilities for batch parallelization of SGD are rather limited. Despite fast convergence, ellipsoid method can result in a greater total number of calls to oracle than SGD, which works decently with small batches. Complexity is quadratic in dimension of the problem, hence the method is suitable for relatively small dimensionalities.

The paper studies a multidimensional convection-diffusion equation with variable coefficients and a nonclassical boundary condition. Two cases are considered: in the first case, the first boundary condition contains the integral of the unknown function with respect to the integration variable $x_\alpha^{}$, and in the second case, the integral of the unknown function with respect to the integration variable $\tau$, denoting the memory effect. Similar problems arise when studying the transport of impurities along the riverbed. For an approximate solution of the problem posed, a locally one-dimensional difference scheme by A.A. Samarskii with order of approximation $O(h^2+\tau)$. In view of the fact that the equation contains the first derivative of the unknown function with respect to the spatial variable $x_\alpha^{}$, the wellknown method proposed by A.A. Samarskii in constructing a monotonic scheme of the second order of accuracy in $h_\alpha^{}$ for a general parabolic type equation containing one-sided derivatives taking into account the sign of $r_\alpha^{}(x,t)$. To increase the boundary conditions of the third kind to the second order of accuracy in $h_\alpha^{}$, we used the equation, on the assumption that it is also valid at the boundaries. The study of the uniqueness and stability of the solution was carried out using the method of energy inequalities. A priori estimates are obtained for the solution of the difference problem in the $L_2^{}$-norm, which implies the uniqueness of the solution, the continuous and uniform dependence of the solution of the difference problem on the input data, and the convergence of the solution of the locally onedimensional difference scheme to the solution of the original differential problem in the $L_2^{}$-norm with speed equal to the order of approximation of the difference scheme. For a two-dimensional problem, a numerical solution algorithm is constructed.

The paper deals with a heat wave motion problem for a degenerate second-order nonlinear parabolic equation with power nonlinearity. The considered boundary condition specifies in a plane the motion equation of the circular zero front of the heat wave. A new numerical-analytical algorithm for solving the problem is proposed. A solution is constructed stepby- step in time using difference time discretization. At each time step, a boundary value problem for the Poisson equation corresponding to the original equation at a fixed time is considered. This problem is, in fact, an inverse Cauchy problem in the domain whose initial boundary is free of boundary conditions and two boundary conditions (Neumann and Dirichlet) are specified on a current boundary (heat wave). A solution of this problem is constructed as the sum of a particular solution to the nonhomogeneous Poisson equation and a solution to the corresponding Laplace equation satisfying the boundary conditions. Since the inhomogeneity depends on the desired function and its derivatives, an iterative solution procedure is used. The particular solution is sought by the collocation method using inhomogeneity expansion in radial basis functions. The inverse Cauchy problem for the Laplace equation is solved by the null field method as applied to a circular domain with a circular hole. This method is used for the first time to solve such problem. The calculation algorithm is optimized by parallelizing the computations. The parallelization of the computations allows us to realize effectively the algorithm on high performance computing servers. The algorithm is implemented as a program, which is parallelized by using the OpenMP standard for the C++ language, suitable for calculations with parallel cycles. The effectiveness of the algorithm and the robustness of the program are tested by the comparison of the calculation results with the known exact solution as well as with the numerical solution obtained earlier by the authors with the use of the boundary element method. The implemented computational experiment shows good convergence of the iteration processes and higher calculation accuracy of the proposed new algorithm than of the previously developed one. The solution analysis allows us to select the radial basis functions which are most suitable for the proposed algorithm.

The work is devoted to the development of a computational algorithm of the Cartesian grid method for studying the interaction of a shock wave with moving bodies with a piecewise linear boundary. The interest in such problems is connected with direct numerical simulation of two-phase media flows. The effect of the particle shape can be important in the problem of dust layer dispersion behind a passing shock wave. Experimental data on the coefficient of aerodynamic drag of non-spherical particles are practically absent.

Mathematical model is based on the two-dimensional Euler equations, which are solved in a region with varying boundaries. The defining system of equations is integrated using an explicit scheme and the Cartesian grid method. The computational algorithm at the time integration step includes: determining the step value, calculating the dynamics of the body movement (determining the force and moment acting on the body; determining the linear and angular velocities of the body; calculating the new coordinates of the body), calculating the gas parameters. At each time step, all cells are divided into two classes – external (inside the body or intersected by its boundaries) and internal (completely filled with gas). The solution of the Euler equations is constructed only in the internal ones. The main difficulty is the calculation of the numerical flux through the edges common to the internal and external cells intersected by the moving boundaries of the bodies. To calculate this flux, we use a two-wave approximation for solving the Riemann problem and the Steger-Warming scheme. A detailed description of the numerical algorithm is presented.

The efficiency of the algorithm is demonstrated on the problem of lifting a cylinder with a base in the form of a circle, ellipse and rectangle behind a passing shock wave. A circular cylinder test was considered in many papers devoted to the immersed boundary methods development. A qualitative and quantitative analysis of the trajectory of the cylinder center mass is carried out on the basis of comparison with the results of simulations presented in eight other works. For a cylinder with a base in the form of an ellipse and a rectangle, a satisfactory agreement was obtained on the dynamics of its movement and rotation in comparison with the available few literary sources. Grid convergence of the results is investigated for the rectangle. It is shown that the relative error of mass conservation law fulfillment decreases with a linear rate.

In this paper, we introduce a new version of Gauss–Newton method for solving a system of nonlinear equations based on ideas of the residual upper bound for a system of nonlinear equations and a quadratic regularization term. The introduced Gauss–Newton method in practice virtually forms the whole parameterized family of the methods solving systems of nonlinear equations and regression problems. The developed family of Gauss–Newton methods completely consists of iterative methods with generalization for cases of non-euclidean normed spaces, including special forms of Levenberg–Marquardt algorithms. The developed methods use the local model based on a parameterized proximal mapping allowing us to use an inexact oracle of «black–box» form with restrictions for the computational precision and computational complexity. We perform an efficiency analysis including global and local convergence for the developed family of methods with an arbitrary oracle in terms of iteration complexity, precision and complexity of both local model and oracle, problem dimensionality. We present global sublinear convergence rates for methods of the proposed family for solving a system of nonlinear equations, consisting of Lipschitz smooth functions. We prove local superlinear convergence under extra natural non-degeneracy assumptions for system of nonlinear functions. We prove both local and global linear convergence for a system of nonlinear equations under Polyak–Lojasiewicz condition for proposed Gauss– Newton methods. Besides theoretical justifications of methods we also consider practical implementation issues. In particular, for conducted experiments we present effective computational schemes for the exact oracle regarding to the dimensionality of a problem. The proposed family of methods unites several existing and frequent in practice Gauss–Newton method modifications, allowing us to construct a flexible and convenient method implementable using standard convex optimization and computational linear algebra techniques.

The paper is devoted to convex-concave saddle point problems where the objective is a sum of a large number of functions. Such problems attract considerable attention of the mathematical community due to the variety of applications in machine learning, including adversarial learning, adversarial attacks and robust reinforcement learning, to name a few. The individual functions in the sum usually represent losses related to examples from a data set. Additionally, the formulation admits a possibly nonsmooth composite term. Such terms often reflect regularization in machine learning problems. We assume that the dimension of one of the variable groups is relatively small (about a hundred or less), and the other one is large. This case arises, for example, when one considers the dual formulation for a minimization problem with a moderate number of constraints. The proposed approach is based on using Vaidya’s cutting plane method to minimize with respect to the outer block of variables. This optimization algorithm is especially effective when the dimension of the problem is not very large. An inexact oracle for Vaidya’s method is calculated via an approximate solution of the inner maximization problem, which is solved by the accelerated variance reduced algorithm Katyusha. Thus, we leverage the structure of the problem to achieve fast convergence. Separate complexity bounds for gradients of different components with respect to different variables are obtained in the study. The proposed approach is imposing very mild assumptions about the objective. In particular, neither strong convexity nor smoothness is required with respect to the low-dimensional variable group. The number of steps of the proposed algorithm as well as the arithmetic complexity of each step explicitly depend on the dimensionality of the outer variable, hence the assumption that it is relatively small.

The paper presents various modifications of the Frank–Wolfe algorithm in the equilibrium traffic assignment problem. The Beckman model is used as a model for experiments. In this article, first of all, attention is paid to the choice of the direction of the basic step of the Frank–Wolfe algorithm. Algorithms will be presented: Conjugate Frank–Wolfe (CFW), Bi-conjugate Frank–Wolfe (BFW), Fukushima Frank –Wolfe (FFW). Each modification corresponds to different approaches to the choice of this direction. Some of these modifications are described in previous works of the authors. In this article, following algorithms will be proposed: N-conjugate Frank–Wolfe (NFW), Weighted Fukushima Frank–Wolfe (WFFW). These algorithms are some ideological continuation of the BFW and FFW algorithms. Thus, if the first algorithm used at each iteration the last two directions of the previous iterations to select the next direction conjugate to them, then the proposed algorithm NFW is using more than $N$ previous directions. In the case of Fukushima Frank–Wolfe, the average of several previous directions is taken as the next direction. According to this algorithm, a modification WFFW is proposed, which uses a exponential smoothing from previous directions. For comparative analysis, experiments with various modifications were carried out on several data sets representing urban structures and taken from publicly available sources. The relative gap value was taken as the quality metric. The experimental results showed the advantage of algorithms using the previous directions for step selection over the classic Frank–Wolfe algorithm. In addition, an improvement in efficiency was revealed when using more than two conjugate directions. For example, on various datasets, the modification 3FW showed the best convergence. In addition, the proposed modification WFFW often overtook FFW and CFW, although performed worse than NFW.

Monte Carlo (MC) reinforcement learning suffers from high sample complexity, especially in environments with sparse rewards, large state spaces, and strongly correlated trajectories that reduce the statistical efficiency of return estimation. These well-known limitations often lead to slow convergence and unstable learning dynamics, particularly in settings where only a small fraction of collected trajectories is actually informative for policy improvement. A key challenge is therefore to identify a compact yet diverse subset of episodes that contributes most to the accuracy of value estimates while preserving sufficient exploration of the environment. To address this challenge, we reformulate episode selection as a Quadratic Unconstrained Binary Optimization (QUBO) problem and solve it using quantum-inspired sampling techniques. Our method, MC+ QUBO, inserts a combinatorial filtering step into the standard MC policy-evaluation pipeline: given a batch of trajectories, it selects a subset that maximizes cumulative reward and encourages broad state-space coverage. This selection procedure is expressed as a QUBO model, where linear terms favor high-return episodes, quadratic terms penalize redundancy between trajectories, and additional coupling terms can be used to enforce coverage-related constraints or promote structural diversity. Within this framework, we investigate two black-box QUBO solvers: Simulated Quantum Annealing (SQA), which emulates tunneling-based exploration of the search landscape, and Simulated Bifurcation (SB), a dynamical-systems-based iterative optimization method. Both solvers demonstrate the ability to efficiently navigate the combinatorial structure of the trajectory-selection problem and to handle batch sizes that are otherwise computationally expensive for exhaustive or deterministic search. Experiments in a finite-horizon GridWorld environment show that MC+QUBO consistently outperforms vanilla MC in convergence speed, stability of return estimates, and final policy quality. These results highlight the promise of quantum-inspired optimization as a practical decision-making subroutine within reinforcement-learning algorithms, offering a scalable way to improve sample efficiency without modifying the underlying learning paradigm.