Abstract
Correlations between variables in complex ecosystems such as weather and financial markets lead to a great amount of dynamic and co-evolving time series data, posing a significant challenge to the current forecast methods. Discovering dynamic patterns (aka regimes) is crucial to an accurate forecast, especially for the interpretability of the outcome. In this paper, we develop a kernel-based method to learn effective representations for capturing dynamically changing regimes. Each such representation accounts for the non-linear interactions among multiple time series, thereby facilitating more effective regime discovery. On the basis of regime information, we build a regression model to forecast all the variables simultaneously for the next multiple time points. The results on six real-life datasets demonstrate that our method can yield the most accurate forecast (with the lowest root mean square error) in comparison with seven predictive models.
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
Ardia, D., Bluteau, K., Boudt, K., Catania, L., Trottier, D.A.: Markov-switching garch models in r: the MSGARCH package. J. Stat. Softw. (2019)
Bazzi, M., Blasques, F., Koopman, S.J., Lucas, A.: Time-varying transition probabilities for Markov regime switching models. J. Time Series Anal. 38, 458–478 (2017)
Box, G.: Box and jenkins: time series analysis, forecasting and control. In: A Very British Affair, pp. 161–215. Springer, Heidelberg (2013). https://doi.org/10.1057/9781137291264_6
Chatigny, P., Chen, R., Patenaude, J.M., Wang, S.: A variable-order regime switching model to identify significant patterns in financial markets. In: ICDM, pp. 887–892. IEEE (2018)
Chen, R., Paschalidis, I.: Selecting optimal decisions via distributionally robust nearest-neighbor regression. Adv. Neural Inf. Process. Syst. 32 (2019)
Dong, X., Dang, B., Zang, H., Li, S., Ma, D.: The prediction trend of enterprise financial risk based on machine learning arima model. J. Theory Pract. Eng. Sci. 4(01), 65–71 (2024)
Elhamifar, E., Vidal, R.: Sparse subspace clustering: algorithm, theory, and applications. IEEE TPAMI 35(11), 2765–2781 (2013)
Hochstein, A., Ahn, H.I., Leung, Y.T., Denesuk, M.: Switching vector autoregressive models with higher-order regime dynamics application to prognostics and health management. In: PHM, pp. 1–10. IEEE (2014)
Lu, C., Feng, J., Lin, Z., Mei, T., Yan, S.: Subspace clustering by block diagonal representation. IEEE TPAMI 41(2), 487–501 (2018)
Matsubara, Y., Sakurai, Y.: Regime shifts in streams: real-time forecasting of co-evolving time sequences. In: ACM SIGKDD, pp. 1045–1054 (2016)
Matsubara, Y., Sakurai, Y.: Dynamic modeling and forecasting of time-evolving data streams. In: ACM SIGKDD, pp. 458–468 (2019)
Oreshkin, B.N., Carpov, D., Chapados, N., Bengio, Y.: N-beats: Neural basis expansion analysis for interpretable time series forecasting. arXiv preprint arXiv:1905.10437 (2019)
Qiao, Y., Jin, J., Ni, F., Yu, J., Chen, W.: Application of machine learning in financial risk early warning and regional prevention and control: a systematic analysis based on shap. World Trends Realities Accompany. Prob. Dev. 331 (2023)
Song, X., et al.: Zeroprompt: streaming acoustic encoders are zero-shot masked lms. arXiv preprint arXiv:2305.10649 (2023)
Xu, C., Yu, J., Chen, W., Xiong, J.: Deep learning in photovoltaic power generation forecasting: Cnn-lstm hybrid neural network exploration and research. In: The 3rd International scientific and practical conference “Technologies in education in schools and universities”, Athens, Greece, 23–26 January 2024, vol. 363, p. 295. International Science Group (2024)
Xu, K., Chen, L., Wang, S.: Data-driven kernel subspace clustering with local manifold preservation. In: 2022 IEEE International Conference on Data Mining Workshops (ICDMW), pp. 876–884. IEEE (2022)
Xu, K., Chen, L., Wang, S.: A multi-view kernel clustering framework for categorical sequences. Expert Syst. Appl. 116637 (2022)
Xu, K., Chen, L., Wang, S., Wang, B.: A self-representation model for robust clustering of categorical sequences. In: U, L.H., Xie, H. (eds.) APWeb-WAIM 2018. LNCS, vol. 11268, pp. 13–23. Springer, Cham (2018). https://doi.org/10.1007/978-3-030-01298-4_2
Ye, J., et al.: Multiplexed oam beams classification via fourier optical convolutional neural network. In: 2023 IEEE Photonics Conference (IPC), pp. 1–2. IEEE (2023)
Ye, J., Solyanik, M., Hu, Z., Dalir, H., Nouri, B.M., Sorger, V.J.: Free-space optical multiplexed orbital angular momentum beam identification system using fourier optical convolutional layer based on 4f system. In: Complex Light and Optical Forces XVII, vol. 12436, pp. 70–80. SPIE (2023)
Zhou, H., et al.: Informer: beyond efficient transformer for long sequence time-series forecasting. In: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 35, pp. 11106–11115 (2021)
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2024 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Xu, K., Chen, L., Patenaude, JM., Wang, S. (2024). Kernel Representation Learning with Dynamic Regime Discovery for Time Series Forecasting. In: Yang, DN., Xie, X., Tseng, V.S., Pei, J., Huang, JW., Lin, J.CW. (eds) Advances in Knowledge Discovery and Data Mining. PAKDD 2024. Lecture Notes in Computer Science(), vol 14650. Springer, Singapore. https://doi.org/10.1007/978-981-97-2266-2_20
Download citation
DOI: https://doi.org/10.1007/978-981-97-2266-2_20
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-97-2265-5
Online ISBN: 978-981-97-2266-2
eBook Packages: Computer ScienceComputer Science (R0)