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).

The paper provides a review of the existing methods of signals’ processing within the conditions of the Rice statistical model applicability. There are considered the principle development directions, the existing limitations and the improvement possibilities concerning the methods of solving the tasks of noise suppression and analyzed signals’ filtration by the example of magnetic-resonance visualization. A conception of a new approach to joint calculation of Rician signal’s both parameters has been developed based on the method of moments in two variants of its implementation. The computer simulation and the comparative analysis of the obtained numerical results have been conducted.

This article explores a method of machine learning based on the theory of random functions. One of the main problems of this method is that decision rule of a model becomes more complicated as the number of training dataset examples increases. The decision rule of the model is the most probable realization of a random function and it's represented as a polynomial with the number of terms equal to the number of training examples. In this article we will show the quick way of the number of training dataset examples reduction and, accordingly, the complexity of the decision rule. Reducing the number of examples of training dataset is due to the search and removal of weak elements that have little effect on the final form of the decision function, and noise sampling elements. For each $(x_i,y_i)$-th element sample was introduced the concept of value, which is expressed by the deviation of the estimated value of the decision function of the model at the point $x_i$, built without the $i$-th element, from the true value $y_i$. Also we show the possibility of indirect using weak elements in the process of training model without increasing the number of terms in the decision function. At the experimental part of the article, we show how changed amount of data affects to the ability of the method of generalizing in the classification task.

The article deals with the development of the noise-reduction algorithm based on anisotropic nonlinear data filtering of computed tomography (CT). Analysis of domestic and foreign literature has shown that the most effective algorithms for noise reduction of CT data use complex methods for analyzing and processing data, such as bilateral, adaptive, three-dimensional and other types of filtrations. However, a combination of such techniques is rarely used in practice due to long processing time per slice. In this regard, it was decided to develop an efficient and fast algorithm for noise-reduction based on simplified bilateral filtration method with three-dimensional data accumulation. The algorithm was developed on C ++11 programming language in Microsoft Visual Studio 2015. The main difference of the developed noise reduction algorithm is the use an improved mathematical model of CT noise, based on the distribution of Poisson and Gauss from the logarithmic value, developed earlier by our team. This allows a more accurate determination of the noise level and, thus, the threshold of data processing. As the result of the noise reduction algorithm, processed CT data with lower noise level were obtained. Visual evaluation of the data showed the increased information content of the processed data, compared to original data, the clarity of the mapping of homogeneous regions, and a significant reduction in noise in processing areas. Assessing the numerical results of the algorithm showed a decrease in the standard deviation (SD) level by more than 6 times in the processed areas, and high rates of the determination coefficient showed that the data were not distorted and changed only due to the removal of noise. Usage of newly developed context dynamic threshold made it possible to decrease SD level on every area of data. The main difference of the developed threshold is its simplicity and speed, achieved by preliminary estimation of the data array and derivation of the threshold values that are put in correspondence with each pixel of the CT. The principle of its work is based on threshold criteria, which fits well both into the developed noise reduction algorithm based on anisotropic nonlinear filtration, and another algorithm of noise-reduction. The algorithm successfully functions as part of the MultiVox workstation and is being prepared for implementation in a single radiological network of the city of Moscow.

A classical for nonlinear dynamics model, Brusselator, is considered, being augmented by addition of a third variable, which plays the role of a fast-diffusing inhibitor. The model is investigated in one-dimensional case in the parametric domain, where two types of diffusive instabilities of system’s homogeneous stationary state are manifested: wave instability, which leads to spontaneous formation of autowaves, and Turing instability, which leads to spontaneous formation of stationary dissipative structures, or Turing structures. It is shown that, due to the subcritical nature of Turing bifurcation, the interaction of two instabilities in this system results in spontaneous formation of stationary dissipative structures already before the passage of Turing bifurcation. In response to different perturbations of spatially uniform stationary state, different stable regimes are manifested in the vicinity of the double bifurcation point in the parametric region under study: both pure regimes, which consist of either stationary or autowave dissipative structures; and mixed regimes, in which different modes dominate in different areas of the computational space. In the considered region of the parametric space, the system is multistable and exhibits high sensitivity to initial noise conditions, which leads to blurring of the boundaries between qualitatively different regimes in the parametric region. At that, even in the area of dominance of mixed modes with prevalence of Turing structures, the establishment of a pure autowave regime has significant probability. In the case of stable mixed regimes, a sufficiently strong local perturbation in the area of the computational space, where autowave mode is manifested, can initiate local formation of new stationary dissipative structures. Local perturbation of the stationary homogeneous state in the parametric region under investidation leads to a qualitatively similar map of established modes, the zone of dominance of pure autowave regimes being expanded with the increase of local perturbation amplitude. In two-dimensional case, mixed regimes turn out to be only transient — upon the appearance of localized Turing structures under the influence of wave regime, they eventually occupy all available space.

In this paper, we consider the problem of restoring the correspondence matrix based on the observations of real correspondences in Moscow. Following the conventional approach [Gasnikov et al., 2013], the transport network is considered as a directed graph whose edges correspond to road sections and the graph vertices correspond to areas that the traffic participants leave or enter. The number of city residents is considered constant. The problem of restoring the correspondence matrix is to calculate all the correspondence from the $i$ area to the $j$ area.

To restore the matrix, we propose to use one of the most popular methods of calculating the correspondence matrix in urban studies — the entropy model. In our work, which is based on the work [Wilson, 1978], we describe the evolutionary justification of the entropy model and the main idea of the transition to solving the problem of entropy-linear programming (ELP) in calculating the correspondence matrix. To solve the ELP problem, it is proposed to pass to the dual problem. In this paper, we describe several numerical optimization methods for solving this problem: the Sinkhorn method and the Accelerated Sinkhorn method. We provide numerical experiments for the following variants of cost functions: a linear cost function and a superposition of the power and logarithmic cost functions. In these functions, the cost is a combination of average time and distance between areas, which depends on the parameters. The correspondence matrix is calculated for multiple sets of parameters and then we calculate the quality of the restored matrix relative to the known correspondence matrix.

We assume that the noise in the restored correspondence matrix is Gaussian, as a result, we use the standard deviation as a quality metric. The article provides an overview of gradient-free optimization methods for solving non-convex problems. Since the number of parameters of the cost function is small, we use the grid search method to find the optimal parameters of the cost function. Thus, the correspondence matrix calculated for each set of parameters and then the quality of the restored matrix is evaluated relative to the known correspondence matrix. Further, according to the minimum residual value for each cost function, we determine for which cost function and at what parameter values the restored matrix best describes real correspondence.

Finding the global minimum of a nonconvex function is one of the key and most difficult problems of the modern optimization. In this paper we consider special classes of nonconvex problems which have a clear and distinct global minimum.

In the first part of the paper we consider two classes of «good» nonconvex functions, which can be bounded below and above by a parabolic function. This class of problems has not been widely studied in the literature, although it is rather interesting from an applied point of view. Moreover, for such problems first-order and higher-order methods may be completely ineffective in finding a global minimum. This is due to the fact that the function may oscillate heavily or may be very noisy. Therefore, our new methods use only zero-order information and are based on grid search. The size and fineness of this grid, and hence the guarantee of convergence speed and oracle complexity, depend on the «goodness» of the problem. In particular, we show that if the function is bounded by fairly close parabolic functions, then the complexity is independent of the dimension of the problem. We show that our new methods converge with a linear convergence rate $\log(1/\varepsilon)$ to a global minimum on the cube.

In the second part of the paper, we consider the nonconvex optimization problem from a different angle. We assume that the target minimizing function is the sum of the convex quadratic problem and a nonconvex «noise» function proportional to the distance to the global solution. Considering functions with such noise assumptions for zero-order methods is new in the literature. For such a problem, we use the classical gradient-free approach with gradient approximation through finite differences. We show how the convergence analysis for our problems can be reduced to the standard analysis for convex optimization problems. In particular, we achieve a linear convergence rate for such problems as well.

Experimental results confirm the efficiency and practical applicability of all the obtained methods.

Gradient-free optimization methods or zeroth-order methods are widely used in training neural networks, reinforcement learning, as well as in industrial tasks where only the values of a function at a point are available (working with non-analytical functions). In particular, the method of error back propagation in PyTorch works exactly on this principle. There is a well-known fact that computer calculations use heuristics of floating-point numbers, and because of this, the problem of finiteness of the mantissa arises.

In this paper, firstly, we reviewed the most popular methods of gradient approximation: Finite forward/central difference (FFD/FCD), Forward/Central wise component (FWC/CWC), Forward/Central randomization on $l_2$ sphere (FSSG2/CFFG2); secondly, we described current theoretical representations of the noise introduced by the inaccuracy of calculating the function at a point: adversarial noise, random noise; thirdly, we conducted a series of experiments on frequently encountered classes of problems, such as quadratic problem, logistic regression, SVM, to try to determine whether the real nature of machine noise corresponds to the existing theory. It turned out that in reality (at least for those classes of problems that were considered in this paper), machine noise turned out to be something between adversarial noise and random, and therefore the current theory about the influence of the mantissa limb on the search for the optimum in gradient-free optimization problems requires some adjustment.

In various applications, i. e., astronomical imaging, electron microscopy, and tomography, images are often damaged by Poisson noise. At the same time, the thermal motion leads to Gaussian noise. Therefore, in such applications, the image is usually corrupted by mixed Poisson – Gaussian noise.

In this paper, we propose a novel method for recovering images corrupted by mixed Poisson – Gaussian noise. In the proposed method, we develop a total variation-based model connected with the nonconvex function and the total generalized variation regularization, which overcomes the staircase artifacts and maintains neat edges.

Numerically, we employ the primal-dual method combined with the classical iteratively reweighted $l_1$ algorithm to solve our minimization problem. Experimental results are provided to demonstrate the superiority of our proposed model and algorithm for mixed Poisson – Gaussian removal to state-of-the-art numerical methods.

Тhe new test-signals forming method for correlation identification of a nonlinear system based on Lee–Shetzen cross-correlation approach is developed and tested. Numerical Gauss–Newton algorithm is applied to correct autocorrelation functions of test signals. The achieved test-signals have length less than 40 000 points and allow to measure the 2nd order Wiener kernels with a linear resolution up to 32 points, the 3rd order Wiener kernels with a linear resolution up to 12 points and the 4th order Wiener kernels with a linear resolution up to 8 points.