Decoupled Dynamic Spatial-Temporal Graph Neural Network for Traffic Forecasting

By Zezhi Shao, Zhao Zhang, Wei Wei, Fei Wang, Yongjun Xu, Xin Cao, Christian S. Jensen. Proceedings of the VLDB Endowment, 2022

378 citations recorded by the source corpus.

We all depend on mobility, and vehicular transportation affects the daily lives of most of us. Thus, the ability to forecast the state of traffic in a road network is an important functionality and a challenging task. Traffic data is often obtained from sensors deployed in a road network. Recent proposals on spatial-temporal graph neural networks have achieved great progress at modeling complex spatial-temporal correlations in traffic data, by modeling traffic data as a diffusion process. However, intuitively, traffic data encompasses two different kinds of hidden time series signals, namely the diffusion signals and inherent signals. Unfortunately, nearly all previous works coarsely consider traffic signals entirely as the outcome of the diffusion, while neglecting the inherent signals, which impacts model performance negatively. To improve modeling performance, we propose a novel Decoupled Spatial-Temporal Framework (DSTF) that separates the diffusion and inherent traffic information in a data-driven manner, which encompasses a unique estimation gate and a residual decomposition mechanism. The separated signals can be handled subsequently by the diffusion and inherent modules separately. Further, we propose an instantiation of DSTF, Decoupled Dynamic Spatial-Temporal Graph Neural Network (D 2 STGNN), that captures spatial-temporal correlations and also features a dynamic graph learning module that targets the learning of the dynamic characteristics of traffic networks. Extensive experiments with four real-world traffic datasets demonstrate that the framework is capable of advancing the state-of-the-art.

Subfields: Building and Construction, Environmental Engineering, Neurology

View Paper: https://www.semanticscholar.org/paper/f005b259c31d1a9ed359a3700e9cbe7f9ca51c0f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Building and Construction

1. Junbo Zhang
2. Yu Zheng
3. Yuxuan Liang
4. Zezhi Shao
5. Fei Wang
6. Zhao Zhang
7. Yongjun Xu
8. Guangyin Jin

In Environmental Engineering

1. Junbo Zhang
2. Yu Zheng

Citations in the corpus

• Spatio-Temporal Graph Neural Networks for Predictive Learning in Urban Computing: A Survey (2023, methodology): “Random initialization-based methods perform adaptive graph structure learning via randomly initialized learnable matrices [35], [57], [59], [91], [158], [159], [160], [161].”
• Spatial-Temporal Identity: A Simple yet Effective Baseline for Multivariate Time Series Forecasting (2022, background): “Recently, Spatial-Temporal Graph Neural Networks (STGNNs) [14, 15, 18] have attracted increasing attention for their state-of-the-art performance.”
• Spatio-Temporal Adaptive Embedding Makes Vanilla Transformer SOTA for Traffic Forecasting (2023, background): “Researchers have devoted considerable effort to developing fancy and complicated models for traffic forecasting, such as novel graph convolutions [2, 3, 8–10, 12, 17, 20, 27, 30, 31, 36, 40], learning graph structures [14, 23, 33, 34, 39], efficient attention mechanisms [5, 19, 32, 42, 43], and othe”