Rethinking spatial-temporal contrastive learning for Urban traffic flow forecasting: multi-level augmentation framework
Abstract Graph neural networks integrating contrastive learning have attracted growing attention in urban traffic flow forecasting. However, most existing graph contrastive learning methods do not perform well in capturing local–global spatial dependencies or designing contrastive learning schemes f...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2024-12-01
|
| Series: | Complex & Intelligent Systems |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s40747-024-01754-z |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832571168635224064 |
|---|---|
| author | Lin Pan Qianqian Ren Zilong Li Xingfeng Lv |
| author_facet | Lin Pan Qianqian Ren Zilong Li Xingfeng Lv |
| author_sort | Lin Pan |
| collection | DOAJ |
| description | Abstract Graph neural networks integrating contrastive learning have attracted growing attention in urban traffic flow forecasting. However, most existing graph contrastive learning methods do not perform well in capturing local–global spatial dependencies or designing contrastive learning schemes for both spatial and temporal dimensions. We argue that these methods can not well extract the spatial-temporal features and are easily affected by data noise. In light of these challenges, this paper proposes an innovative Urban Spatial-Temporal Graph Contrastive Learning framework (UrbanGCL) to improve the accuracy of urban traffic flow forecasting. Specifically, UrbanGCL proposes multi-level data augmentation to address data noise and incompleteness, learn both local and global topology features. The augmented traffic feature matrices and adjacency matrices are then fed into a simple yet effective dual-branch network with shared parameters to capture spatial-temporal correlations within traffic sequences. Moreover, we introduce spatial and temporal contrastive learning auxiliary tasks to alleviate the sparsity of supervision signal and extract the most critical spatial-temporal information. Extensive experimental results on four real-world urban datasets demonstrate that UrbanGCL significantly outperforms other state-of-the-art methods, with the maximum improvement reaching nearly 8.80%. |
| format | Article |
| id | doaj-art-5eb03c0665794b6f80327b2ded2fe0a4 |
| institution | Kabale University |
| issn | 2199-4536 2198-6053 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Springer |
| record_format | Article |
| series | Complex & Intelligent Systems |
| spelling | doaj-art-5eb03c0665794b6f80327b2ded2fe0a42025-02-02T12:49:51ZengSpringerComplex & Intelligent Systems2199-45362198-60532024-12-0111111710.1007/s40747-024-01754-zRethinking spatial-temporal contrastive learning for Urban traffic flow forecasting: multi-level augmentation frameworkLin Pan0Qianqian Ren1Zilong Li2Xingfeng Lv3Department of Computer Science and Technology, Heilongjiang UniversityDepartment of Computer Science and Technology, Heilongjiang UniversityDepartment of Computer Science and Technology, Heilongjiang UniversityDepartment of Computer Science and Technology, Heilongjiang UniversityAbstract Graph neural networks integrating contrastive learning have attracted growing attention in urban traffic flow forecasting. However, most existing graph contrastive learning methods do not perform well in capturing local–global spatial dependencies or designing contrastive learning schemes for both spatial and temporal dimensions. We argue that these methods can not well extract the spatial-temporal features and are easily affected by data noise. In light of these challenges, this paper proposes an innovative Urban Spatial-Temporal Graph Contrastive Learning framework (UrbanGCL) to improve the accuracy of urban traffic flow forecasting. Specifically, UrbanGCL proposes multi-level data augmentation to address data noise and incompleteness, learn both local and global topology features. The augmented traffic feature matrices and adjacency matrices are then fed into a simple yet effective dual-branch network with shared parameters to capture spatial-temporal correlations within traffic sequences. Moreover, we introduce spatial and temporal contrastive learning auxiliary tasks to alleviate the sparsity of supervision signal and extract the most critical spatial-temporal information. Extensive experimental results on four real-world urban datasets demonstrate that UrbanGCL significantly outperforms other state-of-the-art methods, with the maximum improvement reaching nearly 8.80%.https://doi.org/10.1007/s40747-024-01754-zTraffic flow forecastingContrastive learningGraph neural networksSpatial-temporal graphs |
| spellingShingle | Lin Pan Qianqian Ren Zilong Li Xingfeng Lv Rethinking spatial-temporal contrastive learning for Urban traffic flow forecasting: multi-level augmentation framework Complex & Intelligent Systems Traffic flow forecasting Contrastive learning Graph neural networks Spatial-temporal graphs |
| title | Rethinking spatial-temporal contrastive learning for Urban traffic flow forecasting: multi-level augmentation framework |
| title_full | Rethinking spatial-temporal contrastive learning for Urban traffic flow forecasting: multi-level augmentation framework |
| title_fullStr | Rethinking spatial-temporal contrastive learning for Urban traffic flow forecasting: multi-level augmentation framework |
| title_full_unstemmed | Rethinking spatial-temporal contrastive learning for Urban traffic flow forecasting: multi-level augmentation framework |
| title_short | Rethinking spatial-temporal contrastive learning for Urban traffic flow forecasting: multi-level augmentation framework |
| title_sort | rethinking spatial temporal contrastive learning for urban traffic flow forecasting multi level augmentation framework |
| topic | Traffic flow forecasting Contrastive learning Graph neural networks Spatial-temporal graphs |
| url | https://doi.org/10.1007/s40747-024-01754-z |
| work_keys_str_mv | AT linpan rethinkingspatialtemporalcontrastivelearningforurbantrafficflowforecastingmultilevelaugmentationframework AT qianqianren rethinkingspatialtemporalcontrastivelearningforurbantrafficflowforecastingmultilevelaugmentationframework AT zilongli rethinkingspatialtemporalcontrastivelearningforurbantrafficflowforecastingmultilevelaugmentationframework AT xingfenglv rethinkingspatialtemporalcontrastivelearningforurbantrafficflowforecastingmultilevelaugmentationframework |