STAGCN: Spatial–Temporal Attention Graph Convolution Network for Traffic Forecasting
Traffic forecasting plays an important role in intelligent transportation systems. However, the prediction task is highly challenging due to the mixture of global and local spatiotemporal dependencies involved in traffic data. Existing graph neural networks (GNNs) typically capture spatial dependenc...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-05-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/10/9/1599 |
| _version_ | 1827672080400777216 |
|---|---|
| author | Yafeng Gu Li Deng |
| author_facet | Yafeng Gu Li Deng |
| author_sort | Yafeng Gu |
| collection | DOAJ |
| description | Traffic forecasting plays an important role in intelligent transportation systems. However, the prediction task is highly challenging due to the mixture of global and local spatiotemporal dependencies involved in traffic data. Existing graph neural networks (GNNs) typically capture spatial dependencies with the predefined or learnable static graph structure, ignoring the hidden dynamic patterns in traffic networks. Meanwhile, most recurrent neural networks (RNNs) or convolutional neural networks (CNNs) cannot effectively capture temporal correlations, especially for long-term temporal dependencies. In this paper, we propose a spatial–temporal attention graph convolution network (STAGCN), which acquires a static graph and a dynamic graph from data without any prior knowledge. The static graph aims to model global space adaptability, and the dynamic graph is designed to capture local dynamics in the traffic network. A gated temporal attention module is further introduced for long-term temporal dependencies, where a causal-trend attention mechanism is proposed to increase the awareness of causality and local trends in time series. Extensive experiments on four real-world traffic flow datasets demonstrate that STAGCN achieves an outstanding prediction accuracy improvement over existing solutions. |
| first_indexed | 2024-03-10T03:55:43Z |
| format | Article |
| id | doaj.art-8db633be15f542558f0c603048294a51 |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-10T03:55:43Z |
| publishDate | 2022-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-8db633be15f542558f0c603048294a512023-11-23T08:46:42ZengMDPI AGMathematics2227-73902022-05-01109159910.3390/math10091599STAGCN: Spatial–Temporal Attention Graph Convolution Network for Traffic ForecastingYafeng Gu0Li Deng1School of Science, Zhejiang Sci-Tech University, Hangzhou 310018, ChinaSchool of Science, Zhejiang Sci-Tech University, Hangzhou 310018, ChinaTraffic forecasting plays an important role in intelligent transportation systems. However, the prediction task is highly challenging due to the mixture of global and local spatiotemporal dependencies involved in traffic data. Existing graph neural networks (GNNs) typically capture spatial dependencies with the predefined or learnable static graph structure, ignoring the hidden dynamic patterns in traffic networks. Meanwhile, most recurrent neural networks (RNNs) or convolutional neural networks (CNNs) cannot effectively capture temporal correlations, especially for long-term temporal dependencies. In this paper, we propose a spatial–temporal attention graph convolution network (STAGCN), which acquires a static graph and a dynamic graph from data without any prior knowledge. The static graph aims to model global space adaptability, and the dynamic graph is designed to capture local dynamics in the traffic network. A gated temporal attention module is further introduced for long-term temporal dependencies, where a causal-trend attention mechanism is proposed to increase the awareness of causality and local trends in time series. Extensive experiments on four real-world traffic flow datasets demonstrate that STAGCN achieves an outstanding prediction accuracy improvement over existing solutions.https://www.mdpi.com/2227-7390/10/9/1599deep learningtraffic forecastinggraph convolution networksattention mechanismspatial–temporal graph data |
| spellingShingle | Yafeng Gu Li Deng STAGCN: Spatial–Temporal Attention Graph Convolution Network for Traffic Forecasting Mathematics deep learning traffic forecasting graph convolution networks attention mechanism spatial–temporal graph data |
| title | STAGCN: Spatial–Temporal Attention Graph Convolution Network for Traffic Forecasting |
| title_full | STAGCN: Spatial–Temporal Attention Graph Convolution Network for Traffic Forecasting |
| title_fullStr | STAGCN: Spatial–Temporal Attention Graph Convolution Network for Traffic Forecasting |
| title_full_unstemmed | STAGCN: Spatial–Temporal Attention Graph Convolution Network for Traffic Forecasting |
| title_short | STAGCN: Spatial–Temporal Attention Graph Convolution Network for Traffic Forecasting |
| title_sort | stagcn spatial temporal attention graph convolution network for traffic forecasting |
| topic | deep learning traffic forecasting graph convolution networks attention mechanism spatial–temporal graph data |
| url | https://www.mdpi.com/2227-7390/10/9/1599 |
| work_keys_str_mv | AT yafenggu stagcnspatialtemporalattentiongraphconvolutionnetworkfortrafficforecasting AT lideng stagcnspatialtemporalattentiongraphconvolutionnetworkfortrafficforecasting |