Lightweight Design and Implementation of Machine Learning Models in Time Series Forecasting
DOI:
https://doi.org/10.71451/ISTAER2533Keywords:
Time series prediction; Machine learning; Lightweight design; Model compression; Real-time predictionAbstract
With the surge in data volume and the continuous growth of computing requirements, the application of machine learning in time series forecasting faces the challenges of computing resource consumption and real-time requirements. In order to meet this demand, lightweight design has become a key technology to improve the efficiency of time series forecasting models. This paper deeply explores the lightweight design and implementation of machine learning models in time series forecasting, focusing on the application of lightweight technologies such as pruning, quantization, distillation, miniaturized neural networks, and hardware acceleration. By optimizing the network structure and reducing computing resource consumption, the lightweight model can not only improve real-time performance and inference speed, but also ensure high prediction accuracy. Studies have shown that lightweight technology has broad application prospects in fields such as finance, meteorology, and retail. This paper also proposes future research directions for lightweight design, including adaptive lightweight models, the combination of quantum computing and artificial intelligence, and efficient prediction on low-power devices. Finally, this paper looks forward to the optimization and application promotion of lightweight models. It is expected that with the development of technology, lightweight design will be widely used in more emerging fields.
References
[1] Wang, Z., Ruan, S., Huang, T., Zhou, H., Zhang, S., Wang, Y., ... & Liu, Y. (2024). A lightweight multi-layer perceptron for efficient multivariate time series forecasting. Knowledge-Based Systems, 288, 111463. DOI: https://doi.org/10.1016/j.knosys.2024.111463
[2] Ekambaram, V., Jati, A., Nguyen, N., Sinthong, P., & Kalagnanam, J. (2023, August). Tsmixer: Lightweight mlp-mixer model for multivariate time series forecasting. In Proceedings of the 29th ACM SIGKDD conference on knowledge discovery and data mining (pp. 459-469). DOI: https://doi.org/10.1145/3580305.3599533
[3] Coelho, I. M., Coelho, V. N., Luz, E. J. D. S., Ochi, L. S., Guimaraes, F. G., & Rios, E. (2017). A GPU deep learning metaheuristic based model for time series forecasting. Applied Energy, 201, 412-418. DOI: https://doi.org/10.1016/j.apenergy.2017.01.003
[4] Mukherjee, A., Mukherjee, P., Dey, N., De, D., & Panigrahi, B. K. (2020). Lightweight sustainable intelligent load forecasting platform for smart grid applications. Sustainable Computing: Informatics and Systems, 25, 100356. DOI: https://doi.org/10.1016/j.suscom.2019.100356
[5] Hissou, H., Benkirane, S., Guezzaz, A., Azrour, M., & Beni-Hssane, A. (2024). A lightweight time series method for prediction of solar radiation. Energy Systems, 1-38. DOI: https://doi.org/10.1007/s12667-024-00657-9
[6] Torres, J. F., Hadjout, D., Sebaa, A., Martínez-Álvarez, F., & Troncoso, A. (2021). Deep learning for time series forecasting: a survey. Big data, 9(1), 3-21. DOI: https://doi.org/10.1089/big.2020.0159
[7] Campos, D., Zhang, M., Yang, B., Kieu, T., Guo, C., & Jensen, C. S. (2023). LightTS: Lightweight time series classification with adaptive ensemble distillation. Proceedings of the ACM on Management of Data, 1(2), 1-27. DOI: https://doi.org/10.1145/3589316
[8] Lim, B., & Zohren, S. (2021). Time-series forecasting with deep learning: a survey. Philosophical Transactions of the Royal Society A, 379(2194), 20200209. DOI: https://doi.org/10.1098/rsta.2020.0209
[9] Lee, M. C., Lin, J. C., & Katsikas, S. (2024, August). Impact of recurrent neural networks and deep learning frameworks on real-time lightweight time series anomaly detection. In International Conference on Information and Communications Security (pp. 228-247). Singapore: Springer Nature Singapore. DOI: https://doi.org/10.1007/978-981-97-8798-2_12
[10] Chiu, S. M., Chen, Y. C., & Lee, C. (2022). Estate price prediction system based on temporal and spatial features and lightweight deep learning model. Applied Intelligence, 52(1), 808-834. DOI: https://doi.org/10.1007/s10489-021-02472-6
Downloads
Published
Issue
Section
License
Copyright (c) 2025 International Scientific Technical and Economic Research
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
This work is licensed under the Creative Commons Attribution International License (CC BY 4.0).
