Interpolative Bayesian Formulation to Improve Transfer Learning for Anomaly Detection in Rotating Machinery
##plugins.themes.bootstrap3.article.main##
##plugins.themes.bootstrap3.article.sidebar##
Abstract
Anomaly detection in rotating machinery is vital for maintaining industrial reliability, safety, and operational efficiency. However, developing accurate anomaly detection systems remains a significant challenge, particularly in scenarios where labeled anomalous data are limited or entirely absent during training. To address this issue, prior work introduced a transfer learning framework that estimates a key order---a weight vector over features that prioritizes certain dimensions in the anomaly scoring process---from a source domain and applies it to a target domain lacking any anomalous labels. In this study, we extend that framework to more complex and realistic scenarios where the target domain contains a limited number of known anomalous samples. Our enhanced approach estimates the key order within the target domain and integrates it with the source domain’s key order through both Bayesian-based and heuristic methods. We demonstrate that in scenarios with very limited anomalous data in the target, Bayesian-based methods for combining source and target key orders are prone to inaccuracies in estimating the variance of the target key order, leading to suboptimal performance. To mitigate this limitation, we propose a weighted linear combination strategy that improves robustness in regimes with extremely few anomalies. In contrast, when abundant anomalous data are available, the Bayesian-based approach remains preferable due to its capacity to model uncertainty more effectively. Furthermore, we introduce a heuristic based on the Kullback–Leibler (KL) divergence for source domain selection when the optimal source-target pairing is not known \textit{a priori}. Comprehensive experiments conducted on several benchmark datasets for rotating machinery validate the effectiveness of our approach. Results indicate that our proposed method of combining source and target key orders consistently outperforms source-only and target-only key orders across varying levels of anomaly availability. This work underscores the critical role of uncertainty-aware transfer learning and adaptive integration strategies in advancing anomaly detection capabilities in industrial settings characterized by scarce labeling.
##plugins.themes.bootstrap3.article.details##
Machine Learning, Transfer Learning, Bayesian, Rotating Machinery, Anomaly Detection
Karbalayghareh, A., Qian, X., & Dougherty, E. R. (2018). Optimal bayesian transfer learning. IEEE Transactions on Signal Processing, 66(14), 3724–3739.
Kim, M. S., Yun, J. P., & Park, P. (2021). An explainable neural network for fault diagnosis with a frequency activation map. IEEE Access, 9, 98962–98972.
Liang, J., Shui, H., Gupta, R., Upadhyay, D., & Darve, E. (2025). Transfer learning for anomaly detection in rotating machinery using data-driven key order estimation. IEEE Transactions on Automation Science and Engineering, 22, 13310–13326.
Mao, W., Sun, B., & Wang, L. (2021). A new deep dual temporal domain adaptation method for online detection of bearings early fault. Entropy, 23(2), 162.
Mao, W., Wang, G., Kou, L., & Liang, X. (2023). Deep domain-adversarial anomaly detection with one-class transfer learning. IEEE/CAA Journal of Automatica Sinica, 10(2), 524–546.
Mao, W., Zhang, D., Tian, S., & Tang, J. (2020). Robust detection of bearing early fault based on deep transfer learning. Electronics, 9(2), 323.
Mark, W. D., Lee, H., Patrick, R., & Coker, J. D. (2010). A simple frequency-domain algorithm for early detection of damaged gear teeth. Mechanical Systems and Signal Processing, 24(8), 2807–2823.
Mey, O., & Neufeld, D. (2022). Explainable AI algorithms for vibration data-based fault detection: Use case-adapted methods and critical evaluation. Sensors, 22(23), 9037.
Michau, G., & Fink, O. (2021). Unsupervised transfer learning for anomaly detection: Application to complementary operating condition transfer. Knowledge-Based Systems, 216, 106816.
Pan, Q., Bao, Y., & Li, H. (2023). Transfer learning-based data anomaly detection for structural health monitoring. Structural Health Monitoring, 22(5), 3077–3091.
Qin, L. (2024). Rolling bearing fault detection using domain adaptation-based anomaly detection. International Journal on Artificial Intelligence Tools, 33(07), 2440003.
Roelofs, C. M., G¨uck, C., & Faulstich, S. (2024). Transfer learning applications for autoencoder-based anomaly detection in wind turbines. Energy and AI, 17, 100373.
Sarrazin, M., Gillijns, S., Anthonis, J., Janssens, K., van der Auweraer, H., & Verhaeghe, K. (2013). Nvh analysis of a 3 phase 12/8 sr motor drive for hev applications. In 2013 world electric vehicle symposium and exhibition (evs27) (pp. 1–10).
Sharma, V., & Parey, A. (2017). Frequency domain averaging based experimental evaluation of gear fault without tachometer for fluctuating speed conditions. Mechanical Systems and Signal Processing, 85, 278–295.
Shwartz-Ziv, R., Goldblum, M., Souri, H., Kapoor, S., Zhu, C., LeCun, Y., & Wilson, A. G. (2022). Pretrain your loss: Easy bayesian transfer learning with informative priors. Advances in Neural Information Processing Systems, 35, 27706–27715.
Sun, H., Ammann, K., Giannoulakis, S., & Fink, O. (2024). Continuous test-time domain adaptation for efficient fault detection under evolving operating conditions. Proceedings of the European Conference of the PHM Society, 8(1), 11.
Vercruyssen, V., Meert, W., & Davis, J. (2017). Transfer learning for time series anomaly detection. In Proceedings of the workshop and tutorial on interactive adaptive learning@ ecmlpkdd 2017 (Vol. 1924, pp. 27–37).
Vincent, V., Wannes, M., & Jesse, D. (2020). Transfer learning for anomaly detection through localized and unsupervised instance selection. In Proceedings of the aaai conference on artificial intelligence (Vol. 34, pp. 6054–6061).
Wang, B., Baraldi, P., Zio, E., Brown, J., & Gauthier, S. (2026). Fleet-based transfer learning for anomaly detection in industrial systems. Mechanical Systems and Signal Processing, 244, 113725.
Wu, J., Mao, W., Zhang, Y., Fan, L., & Zhong, Z. (2024). A novel few-shot deep transfer learning method for anomaly detection: Deep domain-adversarial contrastive network with time-frequency transferability analytics. IEEE Internet of Things Journal, 11(17), 28809–28823.
Xie, J., Huang, B., & Dubljevic, S. (2021). Transfer learning for dynamic feature extraction using variational bayesian inference. IEEE Transactions on Knowledge and Data Engineering, 34(11), 5524–5535.
Xu, C., Wang, J., Zhang, J., & Li, X. (2021). Anomaly detection of power consumption in yarn spinning using transfer learning. Computers & Industrial Engineering, 152, 107015.
Yadav, E., & Chawla, V. K. (2024). Role and significance of defect detection methods for rotating machines: An explicit literature review. Journal of The Institution of Engineers (India): Series C, 105(5), 1293–1310.
Zhang, S., Ye, F., Wang, B., & Habetler, T. G. (2020). Few-shot bearing anomaly detection via model-agnostic meta-learning. In 2020 23rd international conference on electrical machines and systems (icems) (pp. 1341–1346).
Zhang, S., Zhong, Z., Li, D., Fan, Q., Sun, Y., Zhu, M., . . . others (2022). Efficient kpi anomaly detection through transfer learning for large-scale web services. IEEE Journal on Selected Areas in Communications, 40(8), 2440–2455.
Zhao, Y., Nasrullah, Z., & Li, Z. (2019). Pyod: A python toolbox for scalable outlier detection. Journal of Machine Learning Research, 20(96), 1–7.
Zhu, R., Peng, W., Wang, D., & Huang, C.-G. (2023). Bayesian transfer learning with active querying for intelligent cross-machine fault prognosis under limited data. Mechanical Systems and Signal Processing, 183, 109628.