All issues

Numerous contemporary studies confirm that neurons, astrocytes and blood vessels function as a unified dynamic system. Consequently, the concept of the integrated neurogliovascular unit (NGVU), encompassing these components, has emerged and gained significant traction in recent years. According to this framework, normal brain function relies on a broad complex of interactions between NGVU elements, while the disruption of these links may underlie various neuropathologies. Understanding the processes within a single NGVU, as well as the organization of connections between multiple units, is a prerequisite for successful diagnosis and therapy of neurological disorders.

In this work, we developed an NGVU model that, for the first time, integrates a detailed description of synaptically coupled excitatory and inhibitory neuronal networks (accounting for the E/I balance), extracellular environment dynamics (potassium, glutamate, GABA), and norepinephrine-modulated astrocytic activity, with subsequent regulation of local blood flow.

A key conceptual feature of the model is the integration of multiscale processes — ranging from ion dynamics at the level of individual Hodgkin – Huxley neurons to substance diffusion across a network of 100 NGVUs — into a single system of coupled nonlinear differential equations. This approach enabled the investigation of the ensemble’s collective dynamics and the identification of novel functional regimes.

Numerical experiments established that extracellular potassium dynamics and positive feedback play a decisive role in the formation of stable spatial excitation structures. It is shown that under local stimulation, activity remains confined due to potassium diffusion outflow; however, supercritical excitation initiates self-sustaining autowave regimes. The stabilization of these regimes leads to the formation of spatial patterns morphologically similar to Turing structures. These patterns, characterized by alternating zones of high and low activity, are independent of specific initial conditions but sensitive to parameter variations. This suggests that the system operates in a dynamic instability (chaos) regime, which is consistent with the concept of self-organized criticality of the brain under physiological conditions. The model successfully reproduces experimentally observed phenomena, including bursting and sensitivity to extracellular potassium. The results provide new perspectives for analyzing the pathophysiological mechanisms of brain function.

Storage is the essential and expensive part of cloud computation both from the point of view of network requirements and data access organization. So the choice of storage architecture can be crucial for any application. In this article we can look at the types of cloud architectures for data processing and data storage based on the proven technology of enterprise storage. The advantage of cloud computing is the ability to virtualize and share resources among different applications for better server utilization. We are discussing and evaluating distributed data processing, database architectures for cloud computing and database query in the local network and for real time conditions.

The paper studies the problem of searching for fragments with missing bibliographic links in a scientific article using automatic binary classification. To train the model, we propose a new contrast resampling technique, the innovation of which is the consideration of the context of the link, taking into account the boundaries of the fragment, which mostly affects the probability of presence of a bibliographic links in it. The training set was formed of automatically labeled samples that are fragments of three sentences with class labels «without link» and «with link» that satisfy the requirement of contrast: samples of different classes are distanced in the source text. The feature space was built automatically based on the term occurrence statistics and was expanded by constructing additional features — entities (names, numbers, quotes and abbreviations) recognized in the text.

A series of experiments was carried out on the archives of the scientific journals «Law enforcement review» (273 articles) and «Journal Infectology» (684 articles). The classification was carried out by the models Nearest Neighbors, RBF SVM, Random Forest, Multilayer Perceptron, with the selection of optimal hyperparameters for each classifier.

Experiments have confirmed the hypothesis put forward. The highest accuracy was reached by the neural network classifier (95%), which is however not as fast as the linear one that showed also high accuracy with contrast resampling (91–94%). These values are superior to those reported for NER and Sentiment Analysis on comparable data. The high computational efficiency of the proposed method makes it possible to integrate it into applied systems and to process documents online.

Posts on social networks can both predict the movement of the financial market, and in some cases even determine its direction. The analysis of posts on Twitter contributes to the prediction of cryptocurrency prices. The specificity of the community is represented in a special vocabulary. Thus, slang expressions and abbreviations are used in posts, the presence of which makes it difficult to vectorize text data, as a result of which preprocessing methods such as Stanza lemmatization and the use of regular expressions are considered. This paper describes created simplest machine learning models, which may work despite such problems as lack of data and short prediction timeframe. A word is considered as an element of a binary vector of a data unit in the course of the problem of binary classification solving. Basic words are determined according to the frequency analysis of mentions of a word. The markup is based on Binance candlesticks with variable parameters for a more accurate description of the trend of price changes. The paper introduces metrics that reflect the distribution of words depending on their belonging to a positive or negative classes. To solve the classification problem, we used a dense model with parameters selected by Keras Tuner, logistic regression, a random forest classifier, a naive Bayesian classifier capable of working with a small sample, which is very important for our task, and the k-nearest neighbors method. The constructed models were compared based on the accuracy metric of the predicted labels. During the investigation we recognized that the best approach is to use models which predict price movements of a single coin. Our model deals with posts that mention LUNA project, which no longer exist. This approach to solving binary classification of text data is widely used to predict the price of an asset, the trend of its movement, which is often used in automated trading.

The purpose of the study is to simulate the dynamics of traffic flows on city road networks as well as to systematize the current state of affairs in this area. The introduction states that the development of intelligent transportation systems as an integral part of modern transportation technologies is coming to the fore. The core of these systems contain adequate mathematical models that allow to simulate traffic as close to reality as possible. The necessity of using supercomputers due to the large amount of calculations is also noted, therefore, the creation of special parallel algorithms is needed. The beginning of the article is devoted to the up-to-date classification of traffic flow models and characterization of each class, including their distinctive features and relevant examples with links. Further, the main focus of the article is shifted towards the development of macroscopic and microscopic models, created by the authors, and determination of the place of these models in the aforementioned classification. The macroscopic model is based on the continuum approach and uses the ideology of quasi-gasdynamic systems of equations. Its advantages are indicated in comparison with existing models of this class. The model is presented both in one-dimensional and two-dimensional versions. The both versions feature the ability to study multi-lane traffic. In the two-dimensional version it is made possible by introduction of the concept of “lateral” velocity, i. e., the speed of changing lanes. The latter version allows for carrying out calculations in the computational domain which corresponds to the actual geometry of the road. The section also presents the test results of modeling vehicle dynamics on a road fragment with the local widening and on a road fragment with traffic lights, including several variants of traffic light regimes. In the first case, the calculations allow to draw interesting conclusions about the impact of a road widening on a road capacity as a whole, and in the second case — to select the optimal regime configuration to obtain the “green wave” effect. The microscopic model is based on the cellular automata theory and the single-lane Nagel – Schreckenberg model and is generalized for the multi-lane case by the authors of the article. The model implements various behavioral strategies of drivers. Test computations for the real transport network section in Moscow city center are presented. To achieve an adequate representation of vehicles moving through the network according to road traffic regulations the authors implemented special algorithms adapted for parallel computing. Test calculations were performed on the K-100 supercomputer installed in the Centre of Collective Usage of KIAM RAS.

Epidemic models are widely used to mimic social activity, such as spreading of rumors or panic. Simultaneously, models of excitable media are traditionally used to simulate the propagation of activity. Spreading of activity in the virtual community was simulated within two models: the SIRS epidemic model and the Wiener – Rosenblut model of the excitable media. We used network versions of these models. The network was assumed to be heterogeneous, namely, each element of the network has an individual set of characteristics, which corresponds to different psychological types of community members. The structure of a virtual network relies on an appropriate scale-free network. Modeling was carried out on scale-free networks with various values of the average degree of vertices. Additionally, a special case was considered, namely, a complete graph corresponding to a close professional group, when each member of the group interacts with each. Participants in a virtual community can be in one of three states: 1) potential readiness to accept certain information; 2) active interest to this information; 3) complete indifference to this information. These states correspond to the conditions that are usually used in epidemic models: 1) susceptible to infection, 2) infected, 3) refractory (immune or death due to disease). A comparison of the two models showed their similarity both at the level of main assumptions and at the level of possible modes. Distribution of activity over the network is similar to the spread of infectious diseases. It is shown that activity in virtual networks may experience fluctuations or decay.

Distributed saddle point problems (SPPs) have numerous applications in optimization, matrix games and machine learning. For example, the training of generated adversarial networks is represented as a min-max optimization problem, and training regularized linear models can be reformulated as an SPP as well. This paper studies distributed nonsmooth SPPs with Lipschitz-continuous objective functions. The objective function is represented as a sum of several components that are distributed between groups of computational nodes. The nodes, or agents, exchange information through some communication network that may be centralized or decentralized. A centralized network has a universal information aggregator (a server, or master node) that directly communicates to each of the agents and therefore can coordinate the optimization process. In a decentralized network, all the nodes are equal, the server node is not present, and each agent only communicates to its immediate neighbors.

We assume that each of the nodes locally holds its objective and can compute its value at given points, i. e. has access to zero-order oracle. Zero-order information is used when the gradient of the function is costly, not possible to compute or when the function is not differentiable. For example, in reinforcement learning one needs to generate a trajectory to evaluate the current policy. This policy evaluation process can be interpreted as the computation of the function value. We propose an approach that uses a smoothing technique, i. e., applies a first-order method to the smoothed version of the initial function. It can be shown that the stochastic gradient of the smoothed function can be viewed as a random two-point gradient approximation of the initial function. Smoothing approaches have been studied for distributed zero-order minimization, and our paper generalizes the smoothing technique on SPPs.

The work solves the problem of establishing the dependence of the potential for spatial selection of useful and interfering signals according to the signal-to-interference ratio criterion on the positioning error of user equipment during beamforming by their location at a base station, equipped with an antenna array. Configurable simulation parameters include planar antenna array with a different number of antenna elements, movement trajectory, as well as the accuracy of user equipment location estimation using root mean square error of coordinate estimates. The model implements three algorithms for controlling the shape of the antenna radiation pattern: 1) controlling the beam direction for one maximum and one zero; 2) controlling the shape and width of the main beam; 3) adaptive beamforming. The simulation results showed, that the first algorithm is most effective, when the number of antenna array elements is no more than 5 and the positioning error is no more than 7 m, and the second algorithm is appropriate to employ, when the number of antenna array elements is more than 15 and the positioning error is more than 5 m. Adaptive beamforming is implemented using a training signal and provides optimal spatial selection of useful and interfering signals without device location data, but is characterized by high complexity of hardware implementation. Scripts of the developed models are available for verification. The results obtained can be used in the development of scientifically based recommendations for beam control in ultra-dense millimeter-wave radio access networks of the fifth and subsequent generations.

The paper addresses the problem of modeling the formation and propagation of panic states in social groups with relatively stable structures of interpersonal interactions. Panic is interpreted as a nonlinear process of emotional contagion arising from the interaction between individual psychological characteristics and collective effects within a social environment. In contrast to models focused on the spatial dynamics of moving crowds, the proposed approach concentrates on quasi-stationary interaction networks that reflect informational and emotional contacts among individuals.

The developed discrete network dynamical system integrates individual temperament parameters (sanguine, choleric, phlegmatic, melancholic), the structure of social connections, and nonlinear mechanisms of collective behavior. The individual dynamics of panic are described using an S-shaped growth function, which ensures boundedness of the emotional arousal level and captures the stages of its formation and saturation. Social influence is modeled on a graph of interpersonal interactions (an Erdos –Renyi random network) through local contacts between individuals.

Additionally, the model incorporates the effects of collective contagion and avalanche-like amplification driven by the average panic level in the group, as well as a baseline stress factor depending on group size. Numerical simulation is implemented in a discrete iterative form, allowing for the analysis of both individual and group panic trajectories. A quantitative indicator of the panic propagation rate is introduced, defined by the time required for the group to reach a state close to full panic.

A comparative analysis of heterogeneous and homogeneous groups is conducted, demonstrating that group heterogeneity significantly accelerates panic propagation due to inter-temperament interactions: highly excitable individuals act as initiators of emotional contagion, while more stable individuals partially dampen its dynamics. The evaluation of the model quality using the coefficient of determination shows a high degree of consistency within the simulation data.

The practical significance of the work lies in the potential application of the model for analyzing the resilience of social groups to panic states, assessing risks at mass events, and developing intelligent systems for monitoring collective behavior. Future research directions include extending the model to account for directed and dynamic networks, as well as its calibration based on empirical data.

The paper reviews the main results of the study of real social networks (networks of collaboration between scientists and actors, networks of citation of scientific publications, networks of friends and acquaintances, etc.) and modern online social networks (Twitter, Facebook etc.) from the complex networks theory standpoint. Based on original research by the authors, it reveals peculiarities of perception of certain complex networks.