BAYESIAN APPROACH TO TRAINING DEEP NEURAL NETWORKS FOR RISK MANAGEMENT OF CYBER-PHYSICAL SYSTEMS

Authors

DOI:

https://doi.org/10.28925/2663-4023.2025.28.785

Keywords:

cyber-physical systems; Bayesian approach; neural networks; deep learning models; risk management

Abstract

This article is devoted to the study of the use of the Bayesian approach in training deep neural networks for risk management tasks in cyber-physical systems. Cyber-physical systems (CPS) are complex complexes that combine computational and physical components and require a clear approach to security and risk management. Traditional threat and vulnerability assessment methods often do not take into account the high level of uncertainty inherent in complex environments with the dynamic nature of cyber threats. Instead, the Bayesian approach provides probabilistic modelling and allows for the integration of a priori knowledge of the system into the training of neural networks.

This paper provides a detailed overview of the theoretical foundations of Bayesian Neural Networks (BNNs), compares them with classical deterministic deep learning models, and provides examples of their application in adaptive risk assessment.

The article analyses existing methods of variational Bayesian inference and their role in improving the accuracy and reliability of potential threats forecasting. The main stages of the research methodology, including the design of the neural network architecture.

The results provide experimental evaluations and comparisons with classical approaches, as well as demonstrate practical aspects of implementing Bayesian BNNs for real-world CFAs. The discussion highlights the advantages and disadvantages of the Bayesian approach, emphasises the need to take into account uncertainties and suggests directions for further research. Particular attention is paid to the use of variable quality of sensor data and mechanisms of system response to dynamic conditions in the context of risk assessment.

These conclusions and practical recommendations can serve as a basis for implementing the described methods in various applications, such as industrial automation, intelligent transport systems, critical infrastructure, etc.

Downloads

Download data is not yet available.

References

Lee, E. A. (2008). Cyber Physical Systems: Design Challenges. Proceedings of the 11th IEEE International Symposium on Object-Oriented Real-Time Distributed Computing (ISORC), 363–369. https://doi.org/10.1109/ISORC.2008.25

Baheti, R., Gill, & H. (2011). Cyber-physical systems. The Impact of Control Technology. IEEE Control Systems Society, 161–166. https://ieeecss.org/

Guan, X., Luh, P. B., & Wang, D. (2013). Cyber-physical systems: the next computing revolution for integration of knowledge, control, and services. Proceedings of the ACM/IEEE 4th International Conference on Cyber-Physical Systems (ICCPS), 160–169. https://doi.org/10.1145/2502524.2502530

Cárdenas, A. A., Amin, S., & Sastry, S. (2008). Research challenges for the security of control systems. Proceedings of the 3rd USENIX Workshop on Hot Topics in Security (HotSec), 6–11. https://www.usenix.org/conference/hotsec-08

Neal, R. M. (1996). Bayesian Learning for Neural Networks. Springer. https://doi.org/10.1007/978-1-4612-0745-0.

Blundell, C., Cornebise, J., Kavukcuoglu, K., & Wierstra, D. (2015). Weight Uncertainty in Neural Networks. Proceedings of the 32nd International Conference on Machine Learning (ICML), 1613–1622. http://proceedings.mlr.press/v37/blundell15.html

Kushnir, I., Lesyk, V., & Panasiuk, R. (2021). Secure Communication in Cyber-Physical Systems. Cybersecurity Providing in Information and Telecommunication Systems, 12–28. https://doi.org/10.1007/978-3-030-65722-2_2

Gal, Y., & Ghahramani, Z. (2016). Dropout as a Bayesian approximation: Representing model uncertainty in deep learning. Proceedings of the 33rd International Conference on Machine Learning (ICML), 1050–1059. http://proceedings.mlr.press/v48/gal16.html

Kingma, D. P., & Welling, M. (2013). Auto-Encoding Variational Bayes. arXiv preprint. https://arxiv.org/abs/1312.6114

Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press. http://www.deeplearningbook.org/

Bishop, C. M. (2006). Pattern Recognition and Machine Learning. Springer. https://doi.org/10.1007/978-0-387-45528-0

Abadi, M., Barham, P., & Chen, J., et al. (2016). TensorFlow: A system for large-scale machine learning. Proceedings of the 12th USENIX Symposium on Operating Systems Design and Implementation (OSDI), 265–283. https://www.usenix.org/conference/osdi16/technical-sessions/presentation/abadi

MacKay, D. J. C. (1992). A Practical Bayesian Framework for Backpropagation Networks. Neural Computation, 4(3), 448–472. https://doi.org/10.1162/neco.1992.4.3.448

Graves, A. (). Practical Variational Inference for Neural Networks. Advances in Neural Information Processing Systems (NIPS), 2348–2356. https://proceedings.neurips.cc/paper/2011/hash/7eb3c8be3d41

e8ebfab08eba5f49632-Abstract.html

Jordan, M. I., Ghahramani, Z., Jaakkola, T. S., & Saul, L. K. (1999). An introduction to variational methods for graphical models. Machine Learning, 37(2), 183–233. https://doi.org/10.1023/A:1007665907178

Keras. Deep Learning library for Python. (n. d.). https://keras.io/

Yu, W., Liang, F., & He, X., et al. (2018). A survey on the edge computing for the Internet of Things. IEEE Access, 6, 6900–6919. https://doi.org/10.1109/ACCESS.2017.2778504

Kendall, A., & Gal, Y. (2017). What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision? Advances in Neural Information Processing Systems (NIPS), 5574–5584. https://proceedings.neurips.cc/paper_files/paper/2017/hash/2650d6089a6d640c5e85b2b88265dc2b-Abstract.html

Welling, M., & Teh, Y. W. (2011). Bayesian learning via stochastic gradient Langevin dynamics. Proceedings of the 28th International Conference on Machine Learning (ICML), 681–688. http://proceedings.mlr.press/v28/welling13.html

Qiu, M., & Gai, K. (2020). Reinforcement learning and green communications: The synergy of machine learning and holomorphic encryption. Future Generation Computer Systems, 110, 660–670. https://doi.org/10.1016/j.future.2019.09.017

Wilson, A., & Knowles, D. (2016). Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks. arXiv preprint. https://arxiv.org/abs/1611.00328

Doshi-Velez, F., & Kim, B. (2017). Towards A Rigorous Science of Interpretable Machine Learning. arXiv preprint. https://arxiv.org/abs/1702.08608

Hinton, G. E., & Van Camp, D. (1993). Keeping the neural networks simple by minimizing the description length of the weights. Proceedings of the 6th Annual ACM Conference on Computational Learning Theory (COLT), 5–13. https://doi.org/10.1145/168304.168306

Rybalchenko, L. V., & Gaborets, O. A. (2024). Prokopovych-Tkachenko D.I. Cyber resilience: global threats and national cyber defence strategies. Bulletin of the Academy of Customs Service of Ukraine. Systems and technologies, 2(68), 95–101.

Prokopovych-Tkachenko, D. I., Zverev, V. P., Bushkov, V. G., & Khrushkov, B. S. (2025). Phishing attacks on encrypted messengers: methods, risks and protection recommendations (on the example of Signal messenger). Cybersecurity: Education, Science, Technology, 3(27), 320–328. https://doi.org/10.28925/2663-4023.2025.27.734

Downloads


Abstract views: 73

Published

2025-06-26

How to Cite

Rybalchenko, L., & Prokopovych-Tkachenko, D. (2025). BAYESIAN APPROACH TO TRAINING DEEP NEURAL NETWORKS FOR RISK MANAGEMENT OF CYBER-PHYSICAL SYSTEMS. Electronic Professional Scientific Journal «Cybersecurity: Education, Science, Technique», 4(28), 234–245. https://doi.org/10.28925/2663-4023.2025.28.785