HyperSTLLM: Leveraging Hypergraph Learning for Precise and Scalable Traffic Demand Forecasting with Large Language Models

Welcome to HyperSTLLM's GitHub repository! This project aims to explore the application of hypergraph spatio-temporal learning and Large Language Models (LLMs) in traffic flow demand forecasting. By combining these two advanced techniques, we have developed a novel prediction framework capable of capturing the complex dynamics of transportation networks more precisely and enabling efficient processing of large-scale data.

Our research baselines models refer to the following works and their repository code.

STG4Traffic: {A} Survey and Benchmark of Spatial-Temporal Graph Neural Networks for Traffic Prediction. [Paper][Code].

@article{DBLP:journals/corr/abs-2307-00495,
  author       = {Xunlian Luo and Chunjiang Zhu and Detian Zhang and Qing Li},
  title        = {STG4Traffic: {A} Survey and Benchmark of Spatial-Temporal Graph Neural
                  Networks for Traffic Prediction},
  journal      = {CoRR},
  volume       = {abs/2307.00495},
  year         = {2023}
}

Deep Time Series Models: A Comprehensive Survey and Benchmark. [Paper][Code].

@article{wang2024tssurvey,
  title={Deep Time Series Models: A Comprehensive Survey and Benchmark},
  author={Yuxuan Wang and Haixu Wu and Jiaxiang Dong and Yong Liu and Mingsheng Long and Jianmin Wang},
  booktitle={arXiv preprint arXiv:2407.13278},
  year={2024},
}