I.V. Chebotar1, Yu.N. Gaichuk2, M.A. Sivash3, M.Ya. Mikhlin4

1–4 Military University of Radio Electronics (Cherepovets, Russia)

1–4 vure_nio@mil.ru

This study examines the problem of data-distribution shift in predicting failures of radio-electronic equipment with recurrent neural networks. A typology of distribution-shift classes, their taxonomy, and classical detection methods are presented. Results of statistical tests for normality of the input data and an evaluation of the model’s hidden layer are reported; the analysis demonstrates that structural variations in the hidden layer indicate a shift in the data distribution even when the model’s output error remains unchanged. Finally, the forecasting process and distribution-shift detection are simulated by tracking state dynamics of the output hidden layer of a Long Short-Term Memory recurrent neural network.

Chebotar I.V., Gaichuk Yu.N., Sivash M.A., Mikhlin M.Ya. Modeling the process of failure prediction in radio electronic equipment
using a recurrent neural network. Electromagnetic waves and electronic systems. 2025. V. 30. № 4. P. 86−99. DOI: https://doi.org/10.18127/j15604128-202504-08 (in Russian)

1. Sokolovich M.G., Aksenenko M.A., Zakhariev V.A. Predictive maintenance systems for technological equipment based on artificial intelligence methods. Proceedings of the International Scientific Conference "Information Technologies and Systems 2024".  Minsk: BGUIR. 2024. P. 32–33. (in Russian)
2. Yakovleva M.V., Shalina A.I. Decision-making algorithm for the implementation of predictive equipment maintenance in high-tech companies. Russian Journal of Innovation Economics. 2023. V. 13. № 1. P. 159–172. DOI 10.18334/vinec.13.1.117426. (in Russian)
3. GOST R ISO 13381-1-2016. Condition monitoring and diagnostics of machines. Forecasting the technical condition. Part 1. General guidelines. Moscow: Standartinform. 2017. 20 p. (in Russian)
4. Perlov A.Yu., Pankratov V.A., Timoshenko A.V., Tikhonov V.B. Methodology of data processing of the control and diagnostic system to maintain the energy potential of the long-range radar station. Science Intensive Technologies. 2022. V. 23. № 5. P. 66−73. DOI 10.18127/j19998465-202205-09. (in Russian)
5. Jardine A.K.S., Lin D., Banjevic D. A review on machinery diagnostics and prognostics implementing condition-based maintenance. Mechanical Systems and Signal Processing. 2006. V. 20. № 7. P. 1483–1510. DOI 10.1016/j.ymssp.2005.09.012.
6. BS EN 13306:2017. Maintenance. Maintenance terminology. Brussels: European Committee for Standardization. 2017.
7. Davari N., Veloso B., Costa G.De.A., Pereira P.M., Ribeiro R.P., Gama J. A Survey on Data-Driven Predictive Maintenance for the Railway Industry. Sensors. 2021. V. 21. № 17. P. 5739. DOI 10.3390/s21175739.
8. Le Nguyen M.H. Online machine learning-based predictive maintenance for the railway industry. Machine Learning. Institut Polytech-nique de Paris. 2023. 191 p.
9. Hurtado Ju., Salvati D., Semola R., Bosio M., Lomonaco V. Continual learning for predictive maintenance: Overview and challenges. Intelligent Systems with Applications. 2023. V. 19. P. 200251. DOI 10.1016/j.iswa.2023.200251.
10. Marcus G. Deep learning: A critical appraisal. [Electronic resource] – Access mode: https://arxiv.org/pdf/1801.00631v1, date of reference 12.03.2025.
11. Bayram F., Ahmed B.S., Kassler A. From concept drift to model degradation: An overview on performance-aware drift detectors. Knowledge-Based Systems. 2022. V. 245. P. 108632. DOI 10.1016/j.knosys.2022.108632.
12. Moreno-Torres J.G., Raeder T., Alaiz-Rodríguez R., Chawla N.V., Herrera F. A unifying view on dataset shift in classification. Pattern recognition. 2012. V. 45. № 1. P. 521–530. DOI 10.1016/j.patcog.2011.06.019.
13. Palli A.S., Jaafar Ja., Gomes H.M., Hashmani M.A., Gilal A.R. An Experimental Analysis of Drift Detection Methods on Multi-Class Imbalanced Data Streams. Applied Sciences (Switzerland). 2022. V. 12. № 22. P. 11688. DOI 10.3390/app122211688.
14. Page E.S. Continuous inspection schemes. Biometrika. 1954. V. 41. № 1-2. P. 100.
15. Gama J., Medas P., Castillo G., Rodrigues P. Learning with drift detection. Advances in Artificial Intelligence. 2004. V. 3171. P. 286–295. DOI 10.1007/978-3-540-28645-5_29.
16. Ross G.J., Adams N.M., Tasoulis D.K., Hand D.J. Exponentially weighted moving average charts for detecting concept drift. Pattern Recognition Letters. 2012. V. 33. № 2. P. 191–198. DOI 10.1016/j.patrec.2011.08.019.
17. Sethi T.S., Kantardzic M. On the reliable detection of concept drift from streaming unlabeled data. Expert Systems with Applications. 2017. V. 82. P. 77–99. DOI 10.1016/j.eswa.2017.04.008.
18. Dasu T., Krishnan S., Venkatasubramanian S., Yi K. An information-theoretic approach to detecting changes in multi-dimensional data streams. [Electronic resource] – Access mode: https://cse.hkust.edu.hk/~yike/datadiff/datadiff.pdf, date of reference 12.03.2025.
19. Bifet A., Gavalda R. Learning from time-changing data with adaptive windowing. Proceedings of the 2007 SIAM international conference on data mining. 2007. P. 443–448. DOI 10.1137/1.9781611972771.42.
20. Raab C., Heusinger M., Schleif F.-M. Reactive soft prototype computing for concept drift streams. Neurocomputing. 2020. V. 416. P. 340–351. DOI 10.1016/j.neucom.2019.11.111.
21. Namiot D.E., Ilyushin E.A. Data shift monitoring in machine learning models. International Journal of Open Information Technologies. 2022. V. 10. № 12. P. 84–93. (in Russian)
22. Lu J., Liu A., Dong F., Gu F., Gama J., Zhang G. Learning under Concept Drift: A Review. IEEE Transactions on Knowledge and Data Engineering. 2019. V. 31. № 12. P. 2346–2363. DOI 10.1109/TKDE.2018.2876857.
23. Hochreiter S., Schmidhuber J. Long Short-Term Memory. Neural Computation. 1997. V. 9. № 8. P. 1735–1780. DOI 10.1162/neco. 1997.9.8.1735.
24. Kullback S., Leibler R.A. On Information and Sufficiency. Annals of Mathematical Statistics. 1951. V. 22. № 1. P. 79–86. DOI 10.1214/ aoms/1177729694.
25. Gretton A., Borgwardt K.M., Rasch M.J., Schölkopf B., Smola A.J. A Kernel Two-Sample Test. Journal of Machine Learning Research. 2012. V. 13. P. 723–773.
26. ML Interview Q Series: Implement K-Means clustering from scratch in Python and return cluster labels for each data point. [Electronic resource] – Access mode: https://www.rohan-paul.com/p/ml-interview-q-series-implement-k, date of reference 12.03.2025.
27. The SHAP method for categorical features. [Electronic resource] – Access mode: https://nuancesprog.ru/p/16186/, date of reference 12.03.2025. (in Russian)
28. Berdnikova O.M., Ledeneva T.M. Features of divisive binary clustering in terms of interval values. Bulletin of Voronezh State University. Series: System analysis and Information Technology. 2016. № 4. P. 134–141. (in Russian)
29. Koryachko V.P., Oreshkov V.I. Choosing the number of clusters in clustering problems using the silhouette method. Collection of scientific articles of the IX International Scientific and practical Conference "BIG DATA and high-level analysis". Minsk: Belarusian State University of Informatics and Radioelectronics. 2023. V 1. P. 333–340. (in Russian)
30. Kuznetsov V.O. Using principal component method to analyze supply chains reliability. Logistics and supply chain management. 2018. № 4(87). P. 27–33. (in Russian)
31. Denisenko A.A. Research of dimensionality reduction methods in the pattern recognition problem. International Journal of Applied and Fundamental Research. 2020. № 12. P. 65–69. DOI 10.17513/mjpfi.13161. (in Russian)
32. Weather archive. [Electronic resource] – Access mode: https://arhivpogodi.ru/, date of reference 12.03.2025. (in Russian)
33. Shapiro S.S., Wilk M.B. An analysis of variance test for normality (complete samples). Biometrika. 1965. V. 52. № 3-4. P. 591–611. DOI 10.2307/2333709.
34. Anderson T.W., Darling D.A. Asymptotic theory of certain «goodness of fit» criteria based on stochastic processes. Annals of Mathematical Statistics. 1952. V. 23. № 2. P. 193–212. DOI 10.1214/aoms/1177729437.
35. Report on the implementation of the Interdepartmental Program of Scientific Research and Observations in the Svalbard Archipelago in 2022. [Electronic resource] – Access mode: https://data.aari.ru/rscs_new/files/doc/RSCS_report2022.pdf, date of reference 12.03.2025. (in Russian)
36. Report on the implementation of the Interdepartmental Program of Scientific Research and Observations in the Svalbard Archipelago in 2021. [Electronic resource] – Access mode: https://data.aari.ru/rscs_new/files/doc/RSCS_2021o.pdf, date of reference 12.03.2025. (in Russian)