每個公民都是道路使用者，交通擁堵在大城市中是常見的現象。為了優化道路網絡的使用，預測交通已成為一個重要的議題和熱門的研究領域。研究樣本數據通常是從部署在高速公路網絡上的傳感器獲得的。近來，最先進的交通預測模型通常優先將交通數據視為擴散過程來處理。儘管這些模型被認為能夠理解複雜的空間和時間模式，但它們往往主要集中在設計複雜的擴散模型上，忽略了每一個時刻或傳感器都具有獨特的特性。因此，這些模型依賴於添加組件來增強圖卷積模塊的建模能力，如MixHop和Attention等，以彌補它們在建模獨特性方面的不足。這種對額外組件的依賴通常會導致高計算成本。針對這一忽視，我提出了一種模型——空間時間嵌入門控循環網絡（STEGRN）。主要的改進包括在圖卷積之前加入時間和節點信息，重新設計動態圖形成模塊，以及排除計算較複雜的組件。
在六個真實交通數據集上的全面實驗證明了所提出模型的有效性。儘管其尺寸明顯更緊湊且沒有先前的圖信息，但該模型始終能夠達到與甚至超過之前未考慮每一時刻和傳感器獨特性的最先進模型相當的準確性。這突顯了在建模交通數據時強調每一時刻和傳感器獨特性的重要性。

Every citizen is a road user, and traffic congestion is a common occurrence in major cities. To optimize the utilization of road networks, forecasting traffic has become an important issue and a popular research domain. Researched sample data are usually obtained from sensors deployed in highway networks. Lately, state-of-the-art models for traffic forecasting often prioritize treating traffic data as a diffusion process. While these models are claimed to be powerful to comprehend intricate spatial and temporal patterns, they tend to focus primarily on designing complex diffusion models, but neglect the consideration that each moment or sensor possesses unique properties. As a consequence, these models heavily rely on adding components to enhance modelling capacity for the graph convolution module that alleviate their deficiency in modelling uniqueness, such as MixHop and Attention. This reliance on additional components often leads to high computation costs. In response to this oversight, I propose a model Spatial Temporal Embedded Gated Recurrent Network (STEGRN). The principal enhancements include the incorporation of time and node information preceding the graph convolution, redesigned dynamic graph formation module, along with the exclusion of components with higher computational complexity.
Thorough experimentation conducted across six real-world traffic datasets substantiates the efficacy of the proposed model. Despite its significantly more compact size and the absence of prior graph information, the model consistently achieves accuracy levels comparable to or even surpassing previous state-of-the-art models which did not consider the uniqueness of each moment and sensor. This underscores the importance of emphasizing the uniqueness of each moment and sensor in modelling traffic data.

1. Bai, S.; Kolter, J. Z. and Koltun, V. (2018) "An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling," in ArXiv, vol. abs/1803.01271.
2. Cao, D.; Wang, Y.; Duan, J.; Zhang, C.; Zhu, X.; Huang, C.; Tong, Y.; Xu, B.; Bai, J.; Tong, J. and Zhang, Q. (2020) "Spectral Temporal Graph Neural Network for Multivariate Time-series Forecasting," in Proceedings of the 34th International Conference on Neural Information Processing Systems.
3. Cho, K.; Van Merriënboer, B.; Gulcehre, C.; Bahdanau, D.; Bougares, F., Schwenk, H. and Bengio, Y. (2014) "Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation," Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 1724–1734.
4. Deng, J.; Chen, X.; Jiang, R.; Song, X. and Tsang, I. W. (2021) "ST-Norm: Spatial and Temporal Normalization for Multi-variate Time Series Forecasting," Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pp. 269–278.
5. Dey, R. and Salem, F. M. (2017) "Gate-variants of Gated Recurrent Unit (GRU) Neural Networks," 2017 IEEE 60th International Midwest Symposium on Circuits and Systems (MWSCAS), pp. 1597-1600.
6. Gao, H.; Jiang, R.; Dong, Z.; Deng, J.; Ma, Y. and Song, X. (2023) "Spatio-Temporal-Decoupled Masked Pre-training for Traffic Forecasting," in arXiv preprint arXiv:2312.00516.
7. Hopfield, J. J. (1982) "Neural Networks and Physical Systems with Emergent Collective Computational Abilities," Proceedings of the National Academy of Sciences, vol. 79, no. 8, pp. 2554-2558.
8. Jiang, W.; Luo, J.; He, M. and Gu, W. (2023) "Graph Neural Network for Fraffic Forecasting: The Research Progress," ISPRS International Journal of Geo-Information, vol. 12, no. 3, pp. 100.
9. Jin, G.; Liang, Y.; Fang, Y.; Shao, Z.; Huang, J.; Zhang, J. and Zheng, Y. (2023) "Spatio-temporal Graph Neural Networks for Predictive Learning in Urban Computing: A Survey," in IEEE Transactions on Knowledge and Data Engineering.
10. Ju, W.; Zhao, Y.; Qin, Y.; Yi, S.; Yuan, J.; Xiao, Z.; Luo, X.; Yan, X. and Zhang, M. (2024). “COOL: A Conjoint Perspective on Spatio-Temporal Graph Neural Network for Traffic Forecasting,” Information Fusion, vol. 107, pp. 102341.
11. Kipf, T. N. and Welling, M. (2017) "Semi-Supervised Classification with Graph Convolutional Networks," in International Conference on Learning Representations (ICLR).
12. Lablack, M. and Shen, Y. (2023) “Spatio-temporal Graph Mixformer for Traffic Forecasting,” Expert systems with applications, vol. 228, pp. 120281.
13. Li, F.; Feng, J.; Yan, H.; Jin, G.; Jin, D. and Li, Y. (2023) "Dynamic Graph Convolutional Recurrent Network for Traffic Prediction: Benchmark and Solution," ACM Transactions on Knowledge Discovery from Data, vol. 17, no. 1, pp. 1-21.
14. Li, Y.; Yu, R.; Shahabi, C. and Liu, Y. (2018) "Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting," in International Conference on Learning Representations.
15. Liu, H.; Dong, Z.; Jiang, R.; Deng, J.; Deng, J.; Chen, Q. and Song, X. (2023) "Spatio-Temporal Adaptive Embedding Makes Vanilla Transformer SOTA for Traffic Forecasting," Proceedings of the 32nd ACM International Conference on Information and Knowledge Management, pp. 4125–4129.
16. Luo, Q.; He, S.; Han, X.; Wang, Y. and Li, H. (2024). “LSTTN: A Long-Short Term Transformer-based Spatiotemporal Neural Network for Traffic Flow Forecasting,” Knowledge-Based Systems, vol. 293, pp. 111637.
17. Shao, Z.; Wang, F.; Xu, Y.; Wei, W.; Yu, C.; Zhang, Z.; Yao, D.; Jin, G.; Cao, X.; Cong, G; Jensen, C. S. and Cheng, X. (2023) "Exploring Progress in Multivariate Time Series Forecasting: Comprehensive Benchmarking and Heterogeneity Analysis," in arXiv preprint arXiv:2310.06119.
18. Shao, Z.; Zhang, Z.; Wang, F. and Xu, Y. (2022) "Pre-training Enhanced Spatial-temporal Graph Neural Network for Multivariate Time Series Forecasting," in Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining.
19. Shao, Z.; Zhang, Z.; Wang, F.; Wei, W. and Xu, Y. (2022) "Spatial-Temporal Identity: A Simple yet Effective Baseline for Multivariate Time Series Forecasting," Proceedings of the 31st ACM International Conference on Information & Knowledge Management, pp. 4454–4458.
20. Shao, Z.; Zhang, Z.; Wei, W.; Wang, F.; Xu, Y.; Cao, X., and Jensen, C. S. (2022) "Decoupled Dynamic Spatial-Temporal Graph Neural Network for Traffic Forecasting," Proc. VLDB Endow., vol. 15, no. 11, pp. 2733–2746.
21. Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, L. and Polosukhin, I. (2017) "Attention is All you Need," in Neural Information Processing Systems, 30.
22. Williams, B. M. and Hoel, L. A. (2003) "Modeling and Forecasting Vehicular Traffic Flow as a Seasonal ARIMA Process: Theoretical Basis and Empirical Results," Journal of Transportation Engineering, vol. 129, pp. 664-672.
23. Wu, Z.; Pan, S.; Long, G.; Jiang, J. and Zhang, C. (2019) "Graph WaveNet for Deep Spatial-Temporal Graph Modeling," in International Joint Conference on Artificial Intelligence.
24. Wu, Z.; Pan, S.; Long, G.; Jiang, J.; Chang, X. and Zhang, C. (2020) "Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networks," in Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining.
25. Yu, B.; Yin, H. and Zhu, Z. (2018) "Spatio-Temporal Graph Convolutional Networks: A Deep Learning Framework for Traffic Forecasting," in Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence.
26. Zhao, L.; Song, Y.; Zhang, C.; Liu, Y.; Wang, P.; Lin, T.; Deng, M. and Li, H. (2020) "T-GCN: A Temporal Graph Convolutional Network for Traffic Prediction," IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 9, pp. 3848–3858.