All issues

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.

There were presented the results of the computational and theoretical studies related to assessing of an efficacy and public health impact of a vaccination with a rotavirus vaccine in Ukraine. The required indicators are: the genotype-specific vaccine efficacy, number of the severe illness preventions, hospitalizations, outpatient visits and deaths. The results were obtained in a form of tree of decisions based on Makrov model by using mathematical model with computer simulation. The results showed the significant positive effect of the vaccination compared to no vaccination, in case of high level of vaccine coverage in Ukraine.

Modeling the fight against terrorist, pirate and robbery acts at sea is an urgent scientific task due to the prevalence of force acts and the insufficient number of works on this issue. The actions of pirates and terrorists are diverse. Using a base ship, they can attack ships up to 450–500 miles from the coast. Having chosen the target, they pursue it and use the weapons to board the ship. Actions to free a ship captured by pirates or terrorists include: blocking the ship, predicting where pirates might be on the ship, penetrating (from board to board, by air or from under water) and cleaning up the ship’s premises. An analysis of the special literature on the actions of pirates and terrorists showed that the act of force (and actions to neutralize it) consists of two stages: firstly, blocking the vessel, which consists in forcing it to stop, and secondly, neutralizing the team (terrorist groups, pirates), including penetration of a ship (ship) and its cleaning. The stages of the cycle are matched by indicators — the probability of blocking and the probability of neutralization. The variables of the act of force model are the number of ships (ships, boats) of the attackers and defenders, as well as the strength of the capture group of the attackers and the crew of the ship - the victim of the attack. Model parameters (indicators of naval and combat superiority) were estimated using the maximum likelihood method using an international database of incidents at sea. The values of these parameters are 7.6–8.5. Such high values of superiority parameters reflect the parties' ability to act in force acts. An analytical method for calculating excellence parameters is proposed and statistically substantiated. The following indicators are taken into account in the model: the ability of the parties to detect the enemy, the speed and maneuverability characteristics of the vessels, the height of the vessel and the characteristics of the boarding equipment, the characteristics of weapons and protective equipment, etc. Using the Becker model and the theory of discrete choice, the probability of failure of the force act is estimated. The significance of the obtained models for combating acts of force in the sea space lies in the possibility of quantitative substantiation of measures to protect the ship from pirate and terrorist attacks and deterrence measures aimed at preventing attacks (the presence of armed guards on board the ship, assistance from warships and helicopters).

In neuroscience, the study of odor coding mechanisms requires the analysis of spatial activation patterns of olfactory structures (glomeruli) reconstructed from multiphoton microscopy data. However, the lack of a formal statistical framework for analyzing population-level summary maps limits result reproducibility and hinders the development of predictive models. To address these limitations, we developed a novel methodology for the analysis of olfactory activity maps aggregated across multiple animals, based on the theory of unmarked spatial point processes. The methodology includes a data preprocessing procedure and a numerical analysis algorithm implemented in the R environment using the spatstat package.

The proposed approach enables: (1) transformation of raw glomerular activity maps into point patterns while preserving information about glomerular sizes (replacing size information with local point density is a methodological compromise reflecting the “functional weight” of glomerular input); (2) analysis of point patterns based on spatial morphometric characteristics of domains — regions of stable glomerular activation in the olfactory bulb, each approximated by an ellipse, with ellipse parameters (center coordinates in stereotaxic space, major and minor axis lengths, orientation angles), areas, and intra-ellipse point densities reflecting odorant-specific response signatures; (3) statistical hypothesis testing for spatial randomness (Complete Spatial Randomness) using Ripley’s $K$-function, the nearest-neighbor G-function, and Monte Carlo simulations; (4) synthesis of a parametric pairwise interaction model (Strauss process), whose parameters $(r_{PI}, \gamma)$ have a clear biological interpretation — the spatial interaction scale of glomeruli and the strength of response comodulation, respectively.

The methodology was validated using experimental data obtained from 24 laboratory rats: 10 animals stimulated with camphor and 14 with methyl benzoate. The fitted Strauss models yielded close but odorant-specific parameters: interaction radii $r_{PI}$ of 150 $\mu$m (camphor) and 120 μm (methyl benzoate); interaction parameters $gamma$ of 0.95 and 0.89, respectively. The total domain areas (0.60 mm² and 0.73 mm²) and point densities (38 and 43 points/mm²) calculated at the first stage of analysis are fully consistent with the parametric signatures $(r_{PI}, \gamma)$ of the Strauss model. Model validation using $Q$-$Q$ plots of smoothed residuals confirmed their adequacy.

Our results are consistent with data previously obtained using genetic labeling and functional mapping techniques, demonstrating the correctness of the proposed methodology and the effectiveness of multiphoton laser scanning microscopy for such applications. The proposed framework provides reproducible quantitative assessment of glomerular domains within a unified stereotaxic coordinate system and can be extended to other odorants and biological species. All findings were obtained under anesthesia; extrapolation to active olfactory strategies in awake animals requires further investigation.

The method for mapping of intermediate representations (IR) set of C, C++ programs to vector embedding space is considered to create an empirical estimation framework for static performance prediction using LLVM compiler infrastructure. The usage of embeddings makes programs easier to compare due to avoiding Control Flow Graphs (CFG) and Data Flow Graphs (DFG) direct comparison. This method is based on transformation series of the initial IR such as: instrumentation — injection of artificial instructions in an instrumentation compiler’s pass depending on load offset delta in the current instruction compared to the previous one, mapping of instrumented IR into multidimensional vector with IR2Vec and dimension reduction with t-SNE (t-distributed stochastic neighbor embedding) method. The D1 cache miss ratio measured with perf stat tool is considered as performance metric. A heuristic criterion of programs having more or less cache miss ratio is given. This criterion is based on embeddings of programs in 2D-space. The instrumentation compiler’s pass developed in this work is described: how it generates and injects artificial instructions into IR within the used memory model. The software pipeline that implements the performance estimation based on LLVM compiler infrastructure is given. Computational experiments are performed on synthetic tests which are the sets of programs with the same CFGs but with different sequences of offsets used when accessing the one-dimensional array of a given size. The correlation coefficient between performance metric and distance to the worst program’s embedding is measured and proved to be negative regardless of t-SNE initialization. This fact proves the heuristic criterion to be true. The process of such synthetic tests generation is also considered. Moreover, the variety of performance metric in programs set in such a test is proposed as a metric to be improved with exploration of more tests generators.

Most biomechanical tasks of interest to clinicians can be solved only using personalized mathematical models. Such models allow to formalize and relate key pathophysiological processes, basing on clinically available data evaluate non-measurable parameters that are important for the diagnosis of diseases, predict the result of a therapeutic or surgical intervention. The use of models in clinical practice imposes additional restrictions: clinicians require model validation on clinical cases, the speed and automation of the entire calculated technological chain, from processing input data to obtaining a result. Limitations on the simulation time, determined by the time of making a medical decision (of the order of several minutes), imply the use of reduction methods that correctly describe the processes under study within the framework of reduced models or machine learning tools.

Personalization of models requires patient-oriented parameters, personalized geometry of a computational domain and generation of a computational mesh. Model parameters are estimated by direct measurements, or methods of solving inverse problems, or methods of machine learning. The requirement of personalization imposes severe restrictions on the number of fitted parameters that can be measured under standard clinical conditions. In addition to parameters, the model operates with boundary conditions that must take into account the patient’s characteristics. Methods for setting personalized boundary conditions significantly depend on the clinical setting of the problem and clinical data. Building a personalized computational domain through segmentation of medical images and generation of the computational grid, as a rule, takes a lot of time and effort due to manual or semi-automatic operations. Development of automated methods for setting personalized boundary conditions and segmentation of medical images with the subsequent construction of a computational grid is the key to the widespread use of mathematical modeling in clinical practice.

The aim of this work is to review our solutions for personalization of mathematical models within the framework of three tasks of clinical cardiology: virtual assessment of hemodynamic significance of coronary artery stenosis, calculation of global blood flow after hemodynamic correction of complex heart defects, calculating characteristics of coaptation of reconstructed aortic valve.

Laminar-turbulent transition is the subject of an active research related to improvement of economic efficiency of air vehicles, because in the turbulent boundary layer drag increases, which leads to higher fuel consumption. One of the directions of such research is the search for efficient methods, that can be used to find the position of the transition in space. Using this information about laminar-turbulent transition location when designing an aircraft, engineers can predict its performance and profitability at the initial stages of the project. Traditionally, $e^N$ method is applied to find the coordinates of a laminar-turbulent transition. It is a well known approach in industry. However, despite its widespread use, this method has a number of significant drawbacks, since it relies on parallel flow assumption, which limits the scenarios for its application, and also requires computationally expensive calculations in a wide range of frequencies and wave numbers. Alternatively, flow analysis can be done by using Dynamic Mode Decomposition, which allows one to analyze flow disturbances using flow data directly. Since Dynamic Mode Decomposition is a dimensionality reduction method, the number of computations can be dramatically reduced. Furthermore, usage of Dynamic Mode Decomposition expands the applicability of the whole method, due to the absence of assumptions about the parallel flow in its derivation.

The presented study proposes an approach to finding the location of a laminar-turbulent transition using the Dynamic Mode Decomposition method. The essence of this approach is to divide the boundary layer region into sets of subregions, for each of which the transition point is independently calculated, using Dynamic Mode Decomposition for flow analysis, after which the results are averaged to produce the final result. This approach is validated by laminar-turbulent transition predictions of subsonic and supersonic flows over a 2D flat plate with zero pressure gradient. The results demonstrate the fundamental applicability and high accuracy of the described method in a wide range of conditions. The study focuses on comparison with the $e^N$ method and proves the advantages of the proposed approach. It is shown that usage of Dynamic Mode Decomposition leads to significantly faster execution due to less intensive computations, while the accuracy is comparable to the such of the solution obtained with the $e^N$ method. This indicates the prospects for using the described approach in a real world applications.

The modern digitalization of cities occurs according to laws that are not yet clearly defined. The difficulty in understanding these laws is due to the high uncertainty of urban events and the difficulties in developing adequate analytical models for them. To substantiate management decisions on city digitalization in these conditions, effective methods for processing multidimensional time series are necessary. The objective is to develop a promising method. The proposed method is based on a new model of a neural network control system and algorithms for implementing its main functions. Distinctive features of the model include both structural aspects and new rules for neural network-based substantiation of management decisions on city digitalization, using unsupervised training of the applied recurrent neural network. This model provides for adaptive neural network management of city digitalization according to a sliding program. New rules are proposed for the forward and backward transformation of multidimensional time series, modeling the operation of a streaming recurrent neural network, and identifying feasible programs for city digitalization. By adhering to the proposed rules, it is possible not only to eliminate information loss in such a network but also to ensure its stable operation. The features of this neural network’s operation in substantiating feasible urban digitalization programs are described. The requirements for the software implementation of the proposed method are substantiated. The results of experiments using data for St. Petersburg are presented. These data included monthly volumes of shipped computers, electronic and optical products, as well as monthly volumes of shipped products for all manufacturing industries minus the first-time series values ??for the period from January 2021 to December 2025. The effectiveness of management decisions on the city’s digitalization was predicted based on the total volumes of shipped products to the horizon from January to December 2026. It is shown that the application of the proposed method allows for expanded capabilities for adaptively substantiating feasible management decisions on the city’s digitalization while reducing computer modeling costs.

The article describes a model of a human in the informational economy and demonstrates the multicriteria optimizational approach to the metric analysis of model-generated data. The traditional approach using the identification and study involves the model’s identification by time series and its further prediction. However, this is not possible when some variables are not explicitly observed and only some typical borders or population features are known, which is often the case in the social sciences, making some models pure theoretical. To avoid this problem, we propose a method of metric data analysis (MMDA) for identification and study of such models, based on the construction and analysis of the Kolmogorov – Shannon metric nets of the general population in a multidimensional space of social characteristics. Using this method, the coefficients of the model are identified and the features of its phase trajectories are studied. In this paper, we are describing human according to his role in information processing, considering his awareness and cognitive abilities. We construct two lifetime indices of human capital: creative individual (generalizing cognitive abilities) and productive (generalizing the amount of information mastered by a person) and formulate the problem of their multi-criteria (two-criteria) optimization taking into account life expectancy. This approach allows us to identify and economically justify the new requirements for the education system and the information environment of human existence. It is shown that the Pareto-frontier exists in the optimization problem, and its type depends on the mortality rates: at high life expectancy there is one dominant solution, while for lower life expectancy there are different types of Paretofrontier. In particular, the Pareto-principle applies to Russia: a significant increase in the creative human capital of an individual (summarizing his cognitive abilities) is possible due to a small decrease in the creative human capital (summarizing awareness). It is shown that the increase in life expectancy makes competence approach (focused on the development of cognitive abilities) being optimal, while for low life expectancy the knowledge approach is preferable.