DEVELOPMENT OF DECENTRALIZED FORECASTING SYSTEMS BASED ON MULTI-CLOUD INFRASTRUCTURE

A. Kapiton; T. Franchuk; D. Tyshсhenko; R. Zakharov

doi:10.26906/SUNZ.2026.3.100

Автор(и)

A. Kapiton
T. Franchuk
D. Tyshсhenko
R. Zakharov

DOI:

https://doi.org/10.26906/SUNZ.2026.3.100

Ключові слова:

forecasting systems, cloud infrastructure, distributed systems, data synchronization, network latency, data orchestration, machine learning, computer engineering, information technology

Анотація

Relevance. The study substantiates the need for a cloud environment that ensures a high level of system scalability, enabling the integration of new production facilities, workshops, and equipment without the need for capital investment in proprietary server infrastructure. The subject of the study is the processes of developing and operating decentralized forecasting systems based on multi-cloud infrastructure. The aim of the study is to analyze modern methods for developing decentralized forecasting systems based on multi-cloud infrastructure and to enhance their efficiency, reliability, and fault tolerance. The research objectives are to investigate the architectural aspects of creating decentralized forecasting systems in a cloud environment; to examine the effectiveness of using distributed computing for technical diagnostics; and to determine a software solution for time-series forecasting using recurrent neural networks. to analyze cloud platforms and machine learning libraries for distributed time-series modeling; to identify a key challenge regarding the endto-end analysis of various neural network architectures and formulate technical requirements to overcome this limitation. Research methods. The methodology encompasses systems analysis, mathematical modeling, blockchain technologies, and machine learning methods. Research results. The feasibility of employing cloud and distributed architectures to address diagnostic tasks is substantiated, driven by the specific requirements for processing large volumes of data continuously generated by primary sensors. It is determined that a necessary prerequisite is the development and investigation of a timeseries forecasting software system based neural network models utilizing cloud computing; such a system must implement the complete data processing cycle from acquisition to forecast generation and quality assessment thereby enabling its subsequent application in the field of computer engineering. A structured plan for implementing the system's end-to-end operational cycle is presented. Conclusions. The implementation and deployment of such forecasting systems within a cloud infrastructure, utilizing containerization and orchestration via managed services align with modern engineering standards that ensure high fault tolerance and horizontal scalability for analytical platforms.

Завантажити

Дані для завантаження поки недоступні.

Посилання

1. Hamilton, J. (1994), Time Series Analysis, Princeton University Press, Princeton, doi: https://doi.org/10.2307/j.ctv14jx6sm DOI: https://doi.org/10.2307/j.ctv14jx6sm

2. Chatfield, C. (2005), “Time-series forecasting”, Significance, vol. 2, no. 3, рр. 131–133, doi: https://doi.org/10.1111/j.1740-9713.2005.00117.x DOI: https://doi.org/10.1111/j.1740-9713.2005.00117.x

3. Popovic, D. (2026), Computational Intelligence in Time Series Forecasting, Springer-Verlag, London, doi: https://doi.org/10.1007/1-84628-184-9 DOI: https://doi.org/10.1007/1-84628-184-9

4. Cho, K., Merrienboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., and Bengio, Y. (2014), “Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation”, Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing, Doha, Qatar, pp. 1724–1734, doi: https://doi.org/10.3115/v1/D14-1179 DOI: https://doi.org/10.3115/v1/D14-1179

5. Rashid, T. (2016), Make Your Own Neural Network Scotts Valley: CreateSpace, available at: https://app-01-us-dallastx.cashcowfarmer.com/virtual-library/YsnnFO/8AD266/make-your-own-neural-network__tariq-rashid.pdf

6. James, G., Witten, D., Hastie, T., and Tibshirani R. (2013), An Introduction to Statistical Learning: with Applications in R, Springer, New York, available at: https://www.stat.berkeley.edu/~rabbee/s154/ISLR_First_Printing.pdf

7. Amari, S. (1993), “Backpropagation and stochastic gradient descent method”, Neurocomputing, vol. 5, no. 4-5, pp. 185–196, doi: https://doi.org/10.1016/0925-2312(93)90006-O DOI: https://doi.org/10.1016/0925-2312(93)90006-O

8. Moroz, V., and Milovska, K. (2022), “Wavelet analysis and forecasting of financial time series”, Bulletin of the National Technical University «KhPI». Series: Information and Modeling, vol. 1, no/ 1-2 (7-8), doi: https://doi.org/10.20998/2411-0558.2022.02.11 DOI: https://doi.org/10.20998/2411-0558.2022.02.11

9. Hiransha, M., Gopalakrishnan, E., Menon, V. and Soman, K. (2018), “NSE Stock Market Prediction Using Deep-Learning Models Procedia”, Computer Science, vol. 132, pp. 1351–1362, doi: https://doi.org/10.1016/j.procs.2018.05.050 DOI: https://doi.org/10.1016/j.procs.2018.05.050

10. Bengio, Y., Simard, P., and Frasconi, P. (1994), “Learning long-term dependencies with gradient descent is difficult”, IEEE Transactions on Neural Networks, vol. 5, no. 2, pp. 157–166, doi: https://doi.org/10.1109/72.279181 DOI: https://doi.org/10.1109/72.279181

11. Hochreiter, S., and Schmidhuber, J. (1997), “Long Short-Term Memory”, Neural Computation, vol. 9, no. 8, pp. 1735–1780, doi: https://doi.org/10.1162/neco.1997.9.8.1735 DOI: https://doi.org/10.1162/neco.1997.9.8.1735

12. Chung, J., Gulcehre, C., Cho, K. H., and Bengio, Y. (2014), “Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling”, arXiv, arXiv:1412.3555, doi: https://doi.org/10.48550/arXiv.1412.3555

13. Chung, J., Gulcehre, C., Cho K. H., and Bengio, Y. (2019), “Comparative Analysis of Forecasting Financial Time Series Using ARIMA, LSTM, and BiLSTM”, arXiv, arXiv:1911.09512, doi: https://doi.org/10.48550/arXiv.1911.09512

14. Franchuk, T., Tyshchenko, D., Desiatko, A., and Karpunin, I. (2025), “Features of accounting digitalization processes”, Galician economic journal, vol. 95, no 1, pp. 61–66, doi: https://doi.org/10.33108/galicianvisnyk_tntu2025.01.061 DOI: https://doi.org/10.33108/galicianvisnyk_tntu2025.01.061

15. Kapiton, A., Tyshchenko, D., Franchuk, T., and Desiatko, A. (2025), “Modern Trends in Development and Architectural Features of Computers”, Herald of Khmelnytskyi National University. Technical Sciences, vol. 359(6.1), pp. 230–234, doi: https://doi.org/10.31891/2307-5732-2025-359-32 DOI: https://doi.org/10.31891/2307-5732-2025-359-32

16. Kapiton, A., Tyshchenko, D., Franchuk, T., Desiatko, A., and Svystun M. (2025), “Advantages and disadvantages of iOS 26”, Information technology and society, vol. 4 (19), pp. 74–77, doi: https://doi.org/10.32689/maup.it.2025.4.12 DOI: https://doi.org/10.32689/maup.it.2025.4.12