Home > Research > Publications & Outputs > REFOL

Associated organisational unit

Electronic data

Links

Text available via DOI:

View graph of relations

REFOL: Resource-Efficient Federated Online Learning for Traffic Flow Forecasting

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

REFOL: Resource-Efficient Federated Online Learning for Traffic Flow Forecasting. / Liu, Qingxiang; Sun, Sheng; Liang, Yuxuan et al.
In: IEEE Transactions on Intelligent Transportation Systems, Vol. 26, No. 2, 28.02.2025, p. 2777-2792.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Liu, Q, Sun, S, Liang, Y, Xu, X, Liu, M, Bilal, M, Wang, Y, Li, X & Zheng, Y 2025, 'REFOL: Resource-Efficient Federated Online Learning for Traffic Flow Forecasting', IEEE Transactions on Intelligent Transportation Systems, vol. 26, no. 2, pp. 2777-2792. https://doi.org/10.1109/tits.2024.3510913

APA

Liu, Q., Sun, S., Liang, Y., Xu, X., Liu, M., Bilal, M., Wang, Y., Li, X., & Zheng, Y. (2025). REFOL: Resource-Efficient Federated Online Learning for Traffic Flow Forecasting. IEEE Transactions on Intelligent Transportation Systems, 26(2), 2777-2792. https://doi.org/10.1109/tits.2024.3510913

Vancouver

Liu Q, Sun S, Liang Y, Xu X, Liu M, Bilal M et al. REFOL: Resource-Efficient Federated Online Learning for Traffic Flow Forecasting. IEEE Transactions on Intelligent Transportation Systems. 2025 Feb 28;26(2):2777-2792. Epub 2024 Dec 16. doi: 10.1109/tits.2024.3510913

Author

Liu, Qingxiang ; Sun, Sheng ; Liang, Yuxuan et al. / REFOL : Resource-Efficient Federated Online Learning for Traffic Flow Forecasting. In: IEEE Transactions on Intelligent Transportation Systems. 2025 ; Vol. 26, No. 2. pp. 2777-2792.

Bibtex

@article{a45522b4b5d44914b45574059ee18d56,
title = "REFOL: Resource-Efficient Federated Online Learning for Traffic Flow Forecasting",
abstract = "Multiple federated learning (FL) methods are proposed for traffic flow forecasting (TFF) to avoid heavy-transmission and privacy-leaking concerns resulting from the disclosure of raw data in centralized methods. However, these FL methods adopt offline learning which may yield subpar performance, when concept drift occurs, i.e., distributions of historical and future data vary. Online learning can detect concept drift during model training, thus more applicable to TFF. Nevertheless, the existing federated online learning method for TFF fails to efficiently solve the concept drift problem and causes tremendous computing and communication overhead. Therefore, we propose a novel method named Resource-Efficient Federated Online Learning (REFOL) for TFF, which guarantees prediction performance in a communication-lightweight and computation-efficient way. Specifically, we design a data-driven client participation mechanism to detect the occurrence of concept drift and determine clients{\textquoteright} participation necessity. Subsequently, we propose an adaptive online optimization strategy, which guarantees prediction performance and meanwhile avoids meaningless model updates. Then, a graph convolution-based model aggregation mechanism is designed, aiming to assess participants{\textquoteright} contribution based on spatial correlation without importing extra communication and computing consumption on clients. Finally, we conduct extensive experiments on real-world datasets to demonstrate the superiority of REFOL in terms of prediction improvement and resource economization.",
author = "Qingxiang Liu and Sheng Sun and Yuxuan Liang and Xiaolong Xu and Min Liu and Muhammad Bilal and Yuwei Wang and Xujing Li and Yu Zheng",
year = "2025",
month = feb,
day = "28",
doi = "10.1109/tits.2024.3510913",
language = "English",
volume = "26",
pages = "2777--2792",
journal = "IEEE Transactions on Intelligent Transportation Systems",
issn = "1524-9050",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

RIS

TY - JOUR

T1 - REFOL

T2 - Resource-Efficient Federated Online Learning for Traffic Flow Forecasting

AU - Liu, Qingxiang

AU - Sun, Sheng

AU - Liang, Yuxuan

AU - Xu, Xiaolong

AU - Liu, Min

AU - Bilal, Muhammad

AU - Wang, Yuwei

AU - Li, Xujing

AU - Zheng, Yu

PY - 2025/2/28

Y1 - 2025/2/28

N2 - Multiple federated learning (FL) methods are proposed for traffic flow forecasting (TFF) to avoid heavy-transmission and privacy-leaking concerns resulting from the disclosure of raw data in centralized methods. However, these FL methods adopt offline learning which may yield subpar performance, when concept drift occurs, i.e., distributions of historical and future data vary. Online learning can detect concept drift during model training, thus more applicable to TFF. Nevertheless, the existing federated online learning method for TFF fails to efficiently solve the concept drift problem and causes tremendous computing and communication overhead. Therefore, we propose a novel method named Resource-Efficient Federated Online Learning (REFOL) for TFF, which guarantees prediction performance in a communication-lightweight and computation-efficient way. Specifically, we design a data-driven client participation mechanism to detect the occurrence of concept drift and determine clients’ participation necessity. Subsequently, we propose an adaptive online optimization strategy, which guarantees prediction performance and meanwhile avoids meaningless model updates. Then, a graph convolution-based model aggregation mechanism is designed, aiming to assess participants’ contribution based on spatial correlation without importing extra communication and computing consumption on clients. Finally, we conduct extensive experiments on real-world datasets to demonstrate the superiority of REFOL in terms of prediction improvement and resource economization.

AB - Multiple federated learning (FL) methods are proposed for traffic flow forecasting (TFF) to avoid heavy-transmission and privacy-leaking concerns resulting from the disclosure of raw data in centralized methods. However, these FL methods adopt offline learning which may yield subpar performance, when concept drift occurs, i.e., distributions of historical and future data vary. Online learning can detect concept drift during model training, thus more applicable to TFF. Nevertheless, the existing federated online learning method for TFF fails to efficiently solve the concept drift problem and causes tremendous computing and communication overhead. Therefore, we propose a novel method named Resource-Efficient Federated Online Learning (REFOL) for TFF, which guarantees prediction performance in a communication-lightweight and computation-efficient way. Specifically, we design a data-driven client participation mechanism to detect the occurrence of concept drift and determine clients’ participation necessity. Subsequently, we propose an adaptive online optimization strategy, which guarantees prediction performance and meanwhile avoids meaningless model updates. Then, a graph convolution-based model aggregation mechanism is designed, aiming to assess participants’ contribution based on spatial correlation without importing extra communication and computing consumption on clients. Finally, we conduct extensive experiments on real-world datasets to demonstrate the superiority of REFOL in terms of prediction improvement and resource economization.

U2 - 10.1109/tits.2024.3510913

DO - 10.1109/tits.2024.3510913

M3 - Journal article

VL - 26

SP - 2777

EP - 2792

JO - IEEE Transactions on Intelligent Transportation Systems

JF - IEEE Transactions on Intelligent Transportation Systems

SN - 1524-9050

IS - 2

ER -