Search academic texts

Information × Registration Number 0225U001666, (0122U001694) , R & D reports Title Monitoring, forecasting and prevention of crisis phenomena in complex socio-economic systems popup.stage_title Методи прогнозування складних соціально-економічних систем та їх імплементація у систему IPF: - класичні (ARIMA-based, VARbased), методи штучного інтелекту (RNN-based, Random Forest, Support Vector Regression, K-Nearest Neighbors, CART) та комбіновані (Neuralbased “ARIMAs”, AR-Net, DeepVAR) для ідентифікації, попередження та прогнозування критичних і кризових явищ. Head Soloviov Volodymyr M., д.ф.-м.н. Registration Date 04-02-2025 Organization Kryvyi Rih State Pedagogical University popup.description1 The main purpose of the study is to substantiate the conceptual basis of analysis of structural and dynamic properties of complex systems of various natures, to develop theoretical and methodological basis for interdisciplinary tools and economic and mathematical methods of analysis of topological and spectral characteristics, chaotic behavior, evolutionary processes and self-organization, including network-like systems. popup.description2  Financial markets are complex and self-organizing systems that can be viewed as a complex network of market agents in most cases. This project proved that the stock, energy, and cryptocurrency market indices are notable representatives of unbalanced systems characterized by fractal behavior, long memory, and “heavy” tails of the distributions of fluctuations in these market indices, which indicates their nonlinearity and deviation from the normal Gaussian distribution. At the same time, there is currently no understanding of the mechanisms of collective self-organized behavior of economic agents, and crisis phenomena in financial markets are unexpected and hardly predictable. The economic key to coordinating crash events in these markets seems to be the construction of effective indicators-precursors. This project proposed to use approaches based on graph theory, multifractals, fuzzy time series, recurrence analysis, and deep learning methods to monitor the state of the financial system under study and build indicators of potential crashes. The results indicate the potential of using complex systems theory to predict financial collapses, making this approach relevant in the context of risk management, which also indicates their applicability to public opinion and trend analysis. Product Description popup.authors Bielinskyi Andrii O. Mintii Iryna S. Matviichuk Andrii V. Soloviov Volodymyr M. Soloviova Viktoriia V. popup.nrat_date 2025-02-04 Close