Svetlana Sergeevna Mikhaylova

The paper asks what happened to the systemic risk of the Russian stock market after the rupture of external channels in 2022: did it decline together with observable integration, or did it relocate within the domestic market itself? Directional volatility connectedness in the Diebold-Yilmaz framework is estimated under a single protocol on two panels: 15 world stock indices over 19.09.2007–30.06.2026 (4,893 trading days) and 7 sectoral indices of the Moscow Exchange over 09.01.2008–30.06.2026 (4,616 days). External connectedness falls to its historical minimum by 2023–2026 (total index 56.7% against 68.6% in 2009–2019) and the TO and FROM measures of the RTS index shrink to 2.9% and 16.9%, while internal sectoral connectedness settles above its pre-crisis norm (67.4% against 56.2%) and the average inter-sectoral correlation rises from 0.51 to 0.62. The spread of the two levels changes sign for the first time in the sample; endogenous dating places the shift at 01.02.2022, and a unified risk-orientation index, built on a common denominator for internal and external sources, detects the onset of reorientation by late 2021. Systemic risk did not vanish; it changed its geometry, which calls for shifting monitoring emphasis from external indicators to internal cross-sectoral channels.

When modeling the spatial development of a territory, taking into account spatial effects, it is important to keep in mind that the current development of the territory is influenced not only by internal indicators (economic, social, demographic, infrastructural, etc.), but also by the processes taking place in neighboring areas. When modeling the spatial development of the Russian Federation, it is necessary to take into account spatial heterogeneity, long distances, transport corridors and climatic conditions. Accounting for these complex components includes modeling of inter-regional and intra-regional interaction. The aim of the study is to assess the impact of socio-economic factors on the gross regional product (GRP), taking into account the spatial relationship between the federal districts and time dynamics. To achieve the goal, the following tasks were solved in the work: 1) a comprehensive analysis of approaches to modeling the spatial development of regions has been carried out; 2) an adapted methodology of spatial analysis has been developed, including: a comprehensive system of indicators of socio-economic development that takes into account the specifics of Siberian regions, a typology of spatial econometric models. Materials and methods. The econometric spatial modeling apparatus was used in the modeling. Conclusions. Spatial econometric models provide a more accurate description of socio-economic processes in federal districts compared to traditional approaches that do not take into account the spatial structure of data.

The use of mathematical methods to study the dynamics and build forecasts of demographic indicators is possible both with the use of classical econometric models and new machine learning methods. Both approaches have certain advantages and disadvantages and do not allow us to obtain stable parameter estimates and reliable predictive estimates for long-term forecasting. Therefore, the paper proposes to perform a comparative analysis of the econometric approach and machine learning methods in modeling the main demographic indicators of the Russian Federation depending on the source data, which determined the purpose of the work, which is to study the impact of the instability of the source data on the choice of the type of models for long-term forecasting. The research methods were econometric time series models and neural networks. Research results: ARMA models have shown great efficiency for modeling the studied processes. These models have a transparent algorithm for both parameter estimation and their interpretation, make it possible to assess the reliability and significance of parameters, and make interval forecasts with the desired probability, which can be considered as the probability of individual development scenarios.

The article is devoted to the development of a blockchain platform, which is based on the use of smart contract technology and is designed to provide an environment for concluding transactions between agricultural producers and their buyers. Currently, the technology of a distributed registry or Blockchain has received great attention in the field of software development. In many countries, this technology has been introduced in various areas of the economy. There are a large number of companies that have set themselves the goal of understanding the options for using technology to find business development prospects, reduce costs and optimize current processes. Blockchain technology is one of the end-to-end technologies within the framework of the national Digital Economy project.

Education in the modern world is an integral part of the formation of a personality, therefore, it is given special attention. Digital development in the field of higher education requires automation of many university processes, and in order to improve the quality of training specialists and ensure the objectivity of evaluating their achievements, universities are introducing a rating system. The main objectives of such systems are to increase the motivation of students and encourage them to work independently. This paper presents a rating system based on such aspects of student activity as educational, scientific, social, cultural, creative, and sports. The paper uses intellectual analysis of student achievements using the taxonomy of the subject area and machine learning methods. An intelligent system for analyzing student achievements has been developed.

This article presents the results of a study aimed at forecasting signals for buying and selling Bitcoin cryptocurrency using machine learning models. The conducted analysis included the study of cryptocurrency features and markets, technical analysis, development of trading strategies, application of mathematical methods based on moving averages, and building classification models for buy or sell signals. The results demonstrate the effectiveness of applying machine learning models in modern trading strategies in the cryptocurrency market.

Today, solutions to the problem of binary classification using machine learning find applications in a huge number of spheres of life, such as medicine, energy, marketing, agriculture, financial analytics, etc. This is a great opportunity for companies to gain new sources of profit and improve existing processes. Therefore, new solution methods are being actively developed, existing ones are being improved, and research is being conducted on the possibility of using machine learning to solve classification problems in various fields. The study of the effectiveness of using various machine learning methods, taking into account the existing problems of small data in solving the problem of binary classification, is very relevant due to the significant preponderance of developments towards Big Data. For small data, possible problems that affect the effectiveness of the trained model have been identified, and various options for solving these problems have been proposed. To assess the impact of small data problems on the quality of the trained model, a comparative analysis of the quality metrics of models trained on different variations of data processing was carried out. It is concluded that correct work with small data requires timely elimination of such data defects as class imbalance, outliers, etc. In the course of the study, the most significant quality metrics were selected to obtain a model for analyzing medical parameters. A comparative analysis of diabetes detection models based on preprocessed small data has been carried out. For the task under consideration, the stacking model was chosen as the best option for medical use. The results of the analysis showed that machine learning is able to show high efficiency in solving real problems of binary classification.

The article describes a technique for constructing a 3D model of an object based on the resulting images of an object using the Python programming language. As part of the study, an overview of existing solutions and an analysis of the use of algorithms for constructing three-dimensional models were performed. As a result of the work done, software was created that allows you to create a three-dimensional model based on several presented images. The scope of this work is the analysis of an object using a three-dimensional model, as well as the use of three-dimensional terrain models.

Machine learning is the main field of artificial intelligence. This contributes to a new stage in the development of the field of information technology, since now the computer is able to switch to self-learning mode without explicit programming. The aim of the study was to find the optimal set of exogenous variables that ensures the best quality of the model in the task of forecasting output volumes. As a result, several methods of constructing new attributes are implemented and the main aspects in the preprocessing of data from this subject area are highlighted.