A.V. Gulyaev1, S.V. Pivneva2
1, 2 Russian State Social University (Moscow, Russia)

1 Gulyaev81@gmail.com

This scientific article investigates and analyzes the potential application of a complex architecture of bidirectional multilayer neural networks to address the issue of filling missing values in time series data. The problem of missing data is one of the key and critically significant challenges in the context of time series analysis, as the presence of missing values can substantially and negatively impact the reliability and accuracy of models, as well as the adequacy of forecasting. Research in this area highlights that ignoring missing data or using inadequate methods to fill them can lead to distortions in results, which ultimately affects the success and reliability of predictive models and analytical conclusions. Therefore, the development and application of effective methods for handling missing values is an important task to ensure high quality and accuracy in time series modeling. This study proposes a method based on the architecture of bidirectional gated recurrent units, which effectively fills in missing values by taking into account the bidirectional context of the time series. To evaluate the effectiveness of the proposed model, experiments have been conducted on various datasets, including both synthetic and real time series. The results have showed that the proposed approach outperforms traditional methods for filling missing values. The application of the bidirectional neural network model provides a powerful tool for preprocessing temporal data, opening new possibilities for further analysis and modeling of time series.

Gulyaev A.V., Pivneva S.V. Application of bidirectional multilayer neural network for restoring missing values in time series. Neurocomputers. 2025. V. 27. № 2. P. 54–61. DOI: https://doi.org/10.18127/j19998554-202502-06 (in Russian)

1. Gulyaev A.V., Pivneva S.V. Analiz sushchestvuyushchikh regressionnykh, avtoregressionnykh modelej, modelej eksponentsial'nogo sglazhivaniya. Ikh primenenie k prognozirovaniyu pokupatel'skogo sprosa produktov pitaniya, kharakteristiki, osobennosti, dostoinstva i nedostatki. Sb. nauch. trudov Vseross. nauch.-tekhnich. konf. «Nauka. Proizvodstvo. Obrazovanie». M.: Rossijskij gosudarstvennyj sotsial'nyj universitet. 2023. S. 41–51. (in Russian)
2. Kuznetsov R.S., Tumarova T.G., Pilipenko S.V. Prognozirovanie kotirovok aktsij kitajskoj kompanii Nio s ispol'zovaniem dvunapravlennykh LSTM nejronnykh setej. Ekonomika. Pravo. Innovatsii. 2023. № 4. S. 92–101. DOI: 10.17586/2713-1874-2023-4-92-101. (in Russian)
3. Tronov K.A., Fedorov V.O. Rekurrentnye nejronnye seti s upravlyaemymi rekurrentnymi blokami dlya mnogomernykh vremennykh ryadov s propushchennymi znacheniyami. Original'nye issledovaniya. 2022. T. 12. № 12. S. 352–359. (in Russian)
4. Rachit Adhvaryu K.S. Review paper on enhancing COVID-19 fake news detection with transformer model. International Journal on Recent and Innovation Trends in Computing and Communication. 2023. V. 11. № 9. P. 813–824. DOI 10.17762/ijritcc.v11i9.8971.
5. Balti H., Ben Abbes A., Farah I.R. A Bi-GRU-based encoder–decoder framework for multivariate time series forecasting. Soft Computing – A Fusion of Foundations, Methodologies and Applications. 2024. V. 28. P. 6775–6786. DOI: 10.1007/s00500-023-09531-9.
6. Chen Kh., Ignat'eva S., Bogush R., Ablamejko S. Vybor funktsii aktivatsii v svertochnykh nejronnykh setyakh pri povtornoj identifikatsii lyudej v sistemakh videonablyudeniya. Programmirovanie. 2022. № 5. S. 15–26. DOI: 10.31857/S013234742205003X. (in Russian)
7. Nazdryukhin A.S., Fedrak A.M., Radeev N.A. Samonormalizuyushchiesya nejronnye seti s avtomaticheskim podborom giperparametrov dlya klassifikatsii tablichnykh dannykh. Vysokoproizvoditel'nye vychislitel'nye sistemy i tekhnologii. 2021. T. 5. № 1. S. 242–246. (in Russian)
8. Maslennikova A.A., Chernorutskij I.G. Issledovanie i sravnitel'nyj analiz algoritma Adam v modelyakh glubokogo obucheniya. Sb. materialov nauch.-praktich. konf. «Sovremennye tekhnologii v teorii i praktike programmirovaniya». Sankt-Peterburg: FGAOUVO «Sankt-Peterburgskij politekhnicheskij universitet Petra Velikogo». 2021. S. 164–166. (in Russian)
9. Mers M., Yang Zh., Hsieh Yu.An., Tsai Y.Ja. Recurrent neural networks for pavement performance forecasting: Review and model performance comparison. Transportation Research Record. 2023. V. 2677. № 1. P. 610–624. DOI: 10.1177/03611981221100521.
10. Korsakova S.K., Rusina N.V., Bondarik V.M. Prognozirovanie vremennogo ryada v Python s pomoshch'yu metoda Khol'ta-Uintersa. Sb. nauch. statej X Mezhdunarodnoj nauchno-prakticheskoj konferentsii «BIG DATA i analiz vysokogo urovnya». Minsk: Belorusskij gosudarstvennyj universitet informatiki i radioelektroniki. 2024. S. 438–443. (in Russian)
11. Grodzenskij S.Ya., Kalacheva E.A. Informatsionnye tekhnologii: istoriya razvitiya i stanovleniya. Nelinejnyj mir. 2016. № 5. S. 74–79. (in Russian)
12. Grodzenskij S.Ya., Emanakov I.V., Ovchinnikov S.A. Metodika vyyavleniya poter' v proizvodstvennoj sisteme. Naukoemkie tekhnologii. 2018. T. 19. № 1. S. 10–13. (in Russian)
13. Grodzenskij S.Ya. Raspredelenie Vejbulla: istoriya i sovremennost'. Nelinejnyj mir. 2023. T. 21. № 3. S. 54–65. DOI: 10.18127/ j20700970-202303-06. (in Russian)