EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Adaptive Spatio‐Temporal Forecasting With Temporal‐Graph Attention and Adaptive Laplacian Memory for Urban Traffic and Air Quality Prediction

Ali Mohamed, Hamayadji Abdoul Aziz, Alioum Abdoulaye, Francis Yongwa Dtissibe, Ado Adamou Abba Ari, Nabila Labraoui

IET Wireless Sensor Systems · 2026-01 · 16 (1)

DOI: 10.1049/wss2.70043 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IET Wireless Sensor Systems, online erschienen am 18.11.2026. Verfasst von Ali Mohamed und 5 weiteren Autor:innen. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the rapid growth of Internet of things (IoT) applications that incorporate artificial intelligence models into decision‐making processes, the development of federated learning frameworks for intelligent transport systems, environmental monitoring and smart city management has attracted growing interest in recent years, particularly for the prediction of distributed spatiotemporal data. However, many existing approaches still face difficulties in modelling complex nonlinear dependencies and dynamic interactions between spatially distributed regions, particularly due to the use of static aggregation strategies that are ill‐suited to the heterogeneity of the data. This paper proposes an adaptive aggregation strategy based on proximal policy optimisation (PPO), which enables the learning of a dynamic aggregation policy depending on the state of the federated system. At the same time, the LSTM–DSTGCRN model is enhanced by a spatiotemporal attention mechanism with adaptive Laplacian memory (temporal‐graph attention with adaptive Laplacian memory, TGA–ALM). Experiments carried out on the NYC‐Taxi, NYC‐Bike, CHI‐Taxi and Beijing PM2.5 datasets show that the proposed approach improves the accuracy of predictions whilst enhancing the efficiency of the federated learning process compared with baseline methods.

Zitieren

APA

Mohamed, A., Aziz, H. A., Abdoulaye, A., Dtissibe, F. Y., Ari, A. A. A., & Labraoui, N. (2026). Adaptive Spatio‐Temporal Forecasting With Temporal‐Graph Attention and Adaptive Laplacian Memory for Urban Traffic and Air Quality Prediction. IET Wireless Sensor Systems, 16(1). https://doi.org/10.1049/wss2.70043

Vancouver

Mohamed A, Aziz HA, Abdoulaye A, Dtissibe FY, Ari AAA, Labraoui N. Adaptive Spatio‐Temporal Forecasting With Temporal‐Graph Attention and Adaptive Laplacian Memory for Urban Traffic and Air Quality Prediction. IET Wireless Sensor Systems. 2026;16(1). doi:10.1049/wss2.70043

BibTeX
@article{Mohamed2026,
  author = {Mohamed, Ali and Aziz, Hamayadji Abdoul and Abdoulaye, Alioum and Dtissibe, Francis Yongwa and Ari, Ado Adamou Abba and Labraoui, Nabila},
  title = {Adaptive Spatio‐Temporal Forecasting With Temporal‐Graph Attention and Adaptive Laplacian Memory for Urban Traffic and Air Quality Prediction},
  journal = {IET Wireless Sensor Systems},
  year = {2026},
  volume = {16},
  number = {1},
  doi = {10.1049/wss2.70043}
}
Kostenlos starten →