Abstract
In the industrial field, especially the work or environment condition monitoring, it is crucial but difficult to follow the trend of the time series monitoring data (TSD) when the TSD come from different kinds of sensors and are collected by different companies. The privacy of the multi-sensor TSD must be carefully treated. Few studies, however, have been devoted to solving such problems. Federated learning (FL) is a good structure developed by Google for well keeping the personal privacy. Motivated by this, we here present an improved FL structure for not only keeping the data privacy but also extracting and fusing the trends features of the multi-sensor TSD. In our work, the client models of FL are first designed and optimized for getting the initial parameters and features w.r.t. the corresponding sensor’s TSD, and then both the model parameters and the extracted features of all the activated clients (sensors) are sent to the central server and aggregated. The fused parameters and features are returned to the clients and used to update the optimization of the model. Finally, the fused features of all multi-sensor TSD are put into an echo state network (ESN) to fulfill the trend following of the multi-sensor TSD. The proposed algorithm is applied to the multi-sensor electromagnetic radiation intensity TSD sampled from an actual coal mine, and its superiority in promoting the accuracy on every sensor is demonstrated.
This work is supported by the National Natural Science Foundation of China with Grant No. 61876184 and 61473298.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Chen, Y., Hu, Y., Liu, Y.n., Zhu, X.-d.: Processing and fusion for multi-sensor data. J. Jilin Univ. 56(5), 1170–1178 (2018)
Hu, Y., Sun, X., Nie, X., Li, Y., Liu, L.: An enhanced LSTM for trend following of time series. IEEE Access 7, 34020–34030 (2019)
Jaeger, H., Haas, H.: Harnessing nonlinearity: predicting chaotic systems and saving energy in wireless communication. Science 304(5667), 78–80 (2004)
Kamakshi, P.: A survey on privacy issues and privacy preservation in spatial data mining. In: International Conference on Circuit, pp. 1759–1762 (2014)
Konecny, J., Mcmahan, H.B., Ramage, D.: Federated optimization: distributed optimization beyond the datacenter. arXiv: \(\text{Learning}\), pp. 1–5
Konecny, J., Mcmahan, H.B., Ramage, D., Richtrik, P.: Federated optimization: distributed machine learning for on-device intelligence. arXiv preprint arXiv:1610.02527 (2016)
Kumar, A., Kim, J., Lyndon, D., Fulham, M.J., Feng, D.D.: An ensemble of fine-tuned convolutional neural networks for medical image classification. IEEE J. Biomed. Health Inform. 21(1), 31–40 (2017)
Mcmahan, H.B., Moore, E., Ramage, D., Hampson, S., Arcas, B.A.Y.: Communication-efficient learning of deep networks from decentralized data. In: Artifical Intelligence and Statistics, pp. 1273–1282 (2017)
Pan, Q., Wang, Z.f., Liang, Y., Yang, F., Liu, Z.g.: Basic methods and progress of information fusion. Control Theory Appl. 29(10), 1234–1244 (2012)
Putri, A.W., Hira, L.: Hybrid transformation in privacy-preserving data mining. In: International Conference on Data & Software Engineering, pp. 1–6 (2017)
Qi, Y.J., Wang, Q.: Review of multi-source data fusion algorithm. Aerosp. Electron. Warfare 6(33), 37–41 (2017)
Qiu, L., et al.: Characteristics and precursor information of electromagnetic signals of mining-induced coal and gas outburst. J. Loss Prev. Process Ind. 54, 206–215 (2018)
Wang, E.Y., Li, Z.H., He, X.Q., Liang, C.: Application and pre-warning technology of coal and gas outburst by electromagnetic radiation. Coal Sci. Technol. 42(6), 53–57 (2014)
Wen, Y., Wu, J., Yuan, Y.: Multiple-phase modeling of degradation signal for condition monitoring and remaining useful life prediction. IEEE Trans. Reliab. 66(3), 924–938 (2017)
Yao, Q., Tang, J., Zeng, F., Huang, X., Miao, Y., Pan, J.: Feature extraction of SF6 thermal decomposition characteristics to diagnose overheating fault. IET Sci. Meas. Technol. 9(6), 751–757 (2015)
Ye, S., Wu, G.: Recursive annotations for attention-based neural machine translation. In: International Conference on Asian Language Processing (2018)
Zhang, B.C., Lin, J.Q., Chang, Z.C., Yin, X.J., Gao, Z.: The application of multi sensor data fusion based on the improved BP neural network algorithm. In: Control & Decision Conference, pp. 3842–3846 (2016)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Hu, Y., Sun, X., Chen, Y., Lu, Z. (2019). Model and Feature Aggregation Based Federated Learning for Multi-sensor Time Series Trend Following. In: Rojas, I., Joya, G., Catala, A. (eds) Advances in Computational Intelligence. IWANN 2019. Lecture Notes in Computer Science(), vol 11506. Springer, Cham. https://doi.org/10.1007/978-3-030-20521-8_20
Download citation
DOI: https://doi.org/10.1007/978-3-030-20521-8_20
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-20520-1
Online ISBN: 978-3-030-20521-8
eBook Packages: Computer ScienceComputer Science (R0)