Economics and Mathematical Methods

A long-term sustainable development of countries and regions is determined by the nature and quality of the interaction of human economic activity with the environment. Today, at a time of growing shortage of critically needed resources, global and regional anthroposystems are facing a complex environmental crisis. This crisis is also manifesting in profound social and geopolitical changes. The purpose of this work is to substantiate a transdisciplinary approach to modeling flexible and adaptive high-order systems, which allows integrating concepts and data from economic, political, anthropological, environmental and biological research based on systems’ theory. The paper suggests that a synthetic analysis of the interaction of groups of factors different in their characteristics and the dynamics of economic growth of national economies is possible within the framework of an interdisciplinary approach to complex modeling of the behavior of anthropoecosystems, where special attention is paid to the dynamics of their changes, taking into account the degree of their adaptability (provided through the activities of relevant institutions). On the basis of an interdisciplinary approach, a methodology for developing behavioral models of anthropoecosystems was developed, and a set of structural models was proposed that allows studying the interaction of resources, elements and processes in anthropoecomodels, predicting the vector of changes and managing the processes considered in the study, taking into account their complexity.

The main mistakes made by researchers when predicting events using models based on machine learning are discussed. Such errors are: loss of events themselves, due to the construction of abstract features; models are trained on customers rather than events from customers; construction of artificial features; incorrect validation and erroneous model quality metrics; and static parameters are used. An analysis of the mistakes made in one example from Kaggle is provided. The area under the ROC curve for this example is very high — 0.88, but this quality metric is calculated incorrectly. After correcting all errors, the correct metric turned out to be 0.599. A different approach to analyzing and predicting events is presented, which differs significantly from classical machine learning methods. The method is based on consideration of individual mechanisms of event formation for each client. Mechanism models are being built. Using mathematical methods, the parameters of the models of these event formation mechanisms are restored. Parameters are extrapolated to the future. The forecast of a future event is obtained as a result of the functioning of the mechanism model with established parameter values. The model quality metric, the area under the ROC curve, turned out to be 0.615, which is slightly higher than in the Kaggle example, based on machine learning. Thereby, it is shown that the proposed approach is competitive to advanced machine learning techniques.

The article is devoted to the urgent problem of decarbonization of the transport sector at the international level, because transport generates the bulk of greenhouse gas emissions in the 21st century. The aim of the research is to assess the impact of global decarbonization trends on individual types of transport and predict the likely scale of emissions until 2030. Current trends are analyzed taking into account the impact of lockdowns during the COVID-19 pandemic, due to restrictions on the movement of individuals and vehicles, which led to a decrease in the overall level of air pollution and emissions CO₂ and other gases, and post-pandemic times, that demonstrated the restoration and simultaneous reformatting of global supply chains disrupted in the previous years. Developing the provisions of theoretical research on decarbonization issues, based on current statistical data, the authors of the article built predictive models of the dynamics of emissions by transport subsectors. The analysis showed the need to continue the global decarbonization trend, a significant contribution to which can be made by land road transportation (generating the maximum level of pollution for all types of transport) and the increased use of environmental composite components. A shift to greater use of environment-friendly modes of transport (electric vehicles) and sustainable transport technologies will help improve the global climate.

This article is devoted to modeling the salaries of working professions using the methodology of intersectoral balance. The purpose of the study is to identify the main factors influencing the dynamics of remuneration for the specialists in various professions. The authors propose to extend the previously proposed model of price equilibrium of industries in terms of adding working specialties as branches of the economy that have neither final nor intermediate consumption. The effects of changes in the parameters of demand for industrial products from end consumers and the coefficients of direct labor costs are analyzed. As a result of the model experiment, using the example of two industries and two working professions, the following conclusions were drawn: the salary level is closely related to the distribution of employees of various professions by industry; limited labor resources positively affect the wage growth. A decrease in production coefficients for a working specialty, expressed in a decrease in the industry’s need for the labor of these specialists, contrary to popular belief, leads to an increase in wages, i. e. labor becomes more efficient as the production coefficient decreases. The results of the study can be useful to government agencies in forming a strategy for the development of economic sectors, as well as to large companies in controlling personnel policy.

The article examines the problem of assessing the market for small space tugs. The purpose of the article is to analyze the development trends of the commercial segment of the space services’ market associated with the emergence of new space technology, the improvement of space technologies, the need to explore deep space and the problems of attracting funds of private investors. The method of researching assessment of the market for small space tugs is a consistent analysis of the market structure, segments and conditions of use, which is the necessary basis for the study. The analysis of the small space tugs market used in the authors’ study is based on an in-depth analysis of various approaches used in the economics of space activities, and also allows for a visual presentation of results of the data obtained. The article examines both the practical and theoretical studies of the problem, and proposes an approach to ensuring consistency in the process of assessing and analyzing the market for small space tugs. The main trends in the development of space technologies are associated with this final commercial segment. The article analyzes the structure of the small space tugs market, identifies and substantiates approaches to assessing the capacity of small space tugs, taking into account the characteristics of the space market, the search for investment attractiveness and extending their life in orbit. Services provided by small space tugs can be divided into two groups: with a short and extended life cycle. The costs of such services may be taken into account depending on the life cycle of the products.

Human capital is an important driver for sustainable enterprise’s economic growth and becomes more important under digital transformation. The employee profile appears multifaceted due to the expansion of activities. Therefore, the problem of human capital management based on the design of employees’ individual trajectories of professional development is relevant, timely, socially and economically significant. The paper proposes a model for employees’ individual trajectories of the professional development, which is based on reinforcement learning methods. The model forms an optimal management regime and is considered as a consistent set of program activities aimed at the employee’s development in his professional sphere. It considers employee’s individual characteristics (health, competencies, motivation and social capital). The total control system is considered as a digital twin of an employee, and creates the environment — the model of an employee as a Markov decision process and the control model — the agent — a center of enterprise’s decision-making. We use reinforcement learning algorithms DDQN, SARSA, PRO to maximize the agent’s utility function. Based on the experiments, it is shown that the best results are provided by the DDQN algorithm. The results generated by the proposed model are of practical importance, which would contribute to the growth of an enterprise’s innovativeness and competitiveness by improving the human capital quality and increasing the labor resource efficiency.

The study explores the sentiment of Russian private investors in social networks and its impact on the dynamics of the stock return of 78 companies on the Russian stock market (MOEX) in the period from 2018 to 2022. To take into account sentiment when forecasting returns, the authors RSMI index (Russian social media index) is used, which is based on a unique sample of messages from the most popular social networks among Russian investors — “Telegram” and “Tinkoff Pulse”. The RSMI index includes quantitative (the number of publications in relation to each company) and qualitative (investor reactions) characteristics, allowing to determine the real impact of a particular publication on investors. Using the RSMI index, several models for predicting stock prices of Russian companies were used: lasso regression, random forest, gradient boosting, extreme gradient boosting, ensemble learning and long short-term memory. It is demonstrated that for a wide sample of stocks, indicators of technical and fundamental analysis play a large role in building forecasts of changes in stock returns based on hourly data. Although the addition of the sentiment index improves the results of predicting returns for a wide sample of stocks, it does not significantly improve the predictive ability of the models and shows mixed results. The best results of adding the sentiment index to forecast models are shown for the top 15 most discussed Russian companies. For individual models, we achieved an average error reduction of 4.9%, and at the level of specific companies, the MAE error rate was reduced by more than 10% and MSE by 20%. It has been proven that the returns of low-liquidity stocks of the second and third tiers of the Russian stock market are not significantly influenced by the sentiment of private investors on hourly data, and the addition of the sentiment index does not improve the results of forecast models.

The authors propose is formulating a discrete dynamic model of the software development market with aftereffect based on the problem of performance without interruptions in the scheduling theory. In our model unlike the existing problem of performance, interrupts are not allowed. As a result, the problem of performance without interruptions becomes NP-difficult in the case of even two servicing devices, which leads to the need to use the method of branches and boundaries in the resulting discrete dynamic problem with aftereffect in combination with the exact formula for the shortest schedule execution time to determine the lower estimates of the criterion in the intermediate nodes of the search digraph. There is a well-known theorem that in the knapsack problem, the epsilon version of the branches and bounds method is polynomial with a polynomial degree inversely proportional to epsilon. We make an assumption that this is also true for our problem. To test the hypothesis, we conducted the statistical experiment when the parameters of the problem are selected using a random number sensor, and the dimension increases monotonously. The curvature of the graph of the number of exposed vertices from the dimension on a logarithmic scale allows us to estimate the polynomial or exponential nature of the epsilon version of the branch and boundary method in our problem. It is shown that although the approximate algorithm turned out to be exponential, the relative number of revealed vertices decreases very quickly, which shows (reveals) its practical effectiveness.