Trustworthy LLM for traffic forecasting, advances explainable deep and generative models that fuse multimodal sensing with traffic theory to forecast congestion and freight bottlenecks in real time, laying the groundwork for future “traffic copilots” that provide proactive, trustworthy guidance for safer, more efficient, and sustainable mobility systems.
Trustworthy LLM for Traffic Forecasting
Summary
Related Publications
- Guo, X., Zhang, Q., Jiang, J., Peng, M., Zhu, M., Yang, H.F., Towards explainable traffic flow prediction with large language models, Communications in Transportation Research, Volume 4, 2024, 100150, ISSN 2772-4247, https://doi.org/10.1016/j.commtr.2024.100150.
- Zheng, W., Yang, H.F., Cai, J., Wang, P., Jiang, X., Du, S.S., Wang, Y. and Wang, Z., 2023. Integrating the traffic science with representation learning for city-wide network congestion prediction. Information Fusion, 99, p.101837, https://doi.org/10.1016/j.inffus.2023.101837
- Zhang, J., Yang, H.F., Li, A., et al., MLLM-LLaVA-FL: Multimodal Large Language Model Assisted Federated Learning, 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Tucson, AZ, USA, 2025, pp. 4066-4076, https://doi.org/10.1109/WACV61041.2025.00400
External Researchers
Yiran Chen (Duke), Meixin Zhu (HKUST), Simon Du (UW)
Funding Agencies
Collaborating Agencies
