Adaptive Graph Spatial-Temporal Transformer Network for Traffic Forecasting

Feng, Aosong

doi:10.1145/3511808.3557540

Cited by 35 publications

(13 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The third combination is a parallel encoding of the DG by independent transformer and GNN layers, followed by a combination of their encoded hidden states [120], merging the strengths of both layers. Additionally, some dynamic graph models [73], [98], [99], [122], [123], [125], [127] use transformers exclusively as graph encoders, exploiting the self-attention mechanism for node hidden states propagation without relying on traditional GNN architectures.…”

Section: ) Combination Of Transformer With Dgnnsmentioning

confidence: 99%

“…The first method is the embedding based on the time-related index of nodes or edges [98], [100], [120], [124], [127]. The second method utilizes (global) graph spectral information [99], such as the Laplacian or its eigenvectors. The third method relies on spatial domain information or local structures, such as embedding the node degrees or random walks [73], [99], [101], [122], [123], [127], [139].…”

Section: ) Positional Encoding (Pe) On Dynamic Graphsmentioning

confidence: 99%

“…The second method utilizes (global) graph spectral information [99], such as the Laplacian or its eigenvectors. The third method relies on spatial domain information or local structures, such as embedding the node degrees or random walks [73], [99], [101], [122], [123], [127], [139]. Finally, as mentioned in section III-E, almost all the dynamic graph transformers use time embedding to represent the timestamp or snapshot index with a vector.…”

Section: ) Positional Encoding (Pe) On Dynamic Graphsmentioning

confidence: 99%

“…The emergence of Transformer architectures tailored to dynamic graphs marks an evolution towards more diversified temporal encoding methods. Such as in ASTTN [99] which integrates multiple types of positional encoding and learns spatiotemporal attention, as well as in VDGCNeT [101] which combines different attention mechanisms. Meanwhile, the effectiveness of each positional encoding approach and attention mechanisms in different cases and datasets poses an open challenge that warrants further analysis and research.…”

Section: ) Positional Encoding (Pe) On Dynamic Graphsmentioning

confidence: 99%

See 3 more Smart Citations

Dynamic Graph Representation Learning With Neural Networks: A Survey

Yang,

Chatelain,

Adam

2024

IEEE Access

View full text Add to dashboard Cite

In recent years, Dynamic Graph (DG) representations have been increasingly used for modeling dynamic systems due to their ability to integrate both topological and temporal information in a compact representation. Dynamic graphs efficiently handle applications such as social network prediction, recommender systems, traffic forecasting, or electroencephalography analysis, which cannot be addressed using standard numerical representations. As a direct consequence, dynamic graph learning has emerged as a new machine learning problem, combining challenges from both sequential/temporal data processing and static graph learning. In this research area, the Dynamic Graph Neural Network (DGNN) has become the state-of-the-art approach and a plethora of models have been proposed in the very recent years. This paper aims to provide a review of the problems and models related to dynamic graph learning. The various dynamic graph supervised learning settings are analyzed and discussed. We identify the similarities and differences between existing models concerning the way time information is modeled. Finally, we provide guidelines for DGNN design and optimization, and review public datasets for evaluating model performance on various tasks, along with the corresponding publications.

show abstract

Section: ) Combination Of Transformer With Dgnnsmentioning

confidence: 99%

Section: ) Positional Encoding (Pe) On Dynamic Graphsmentioning

confidence: 99%

Section: ) Positional Encoding (Pe) On Dynamic Graphsmentioning

confidence: 99%

Section: ) Positional Encoding (Pe) On Dynamic Graphsmentioning

confidence: 99%

See 2 more Smart Citations

Dynamic Graph Representation Learning With Neural Networks: A Survey

Yang,

Chatelain,

Adam

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Traffic Transformer [27] designs a global encoder and a global-local encoder to extract global spatial features and local spatial features respectively. To model the time delay of spatial information propagation, PDFormer [28] proposes a traffic delay-aware feature transformation module and adaptive STTN [29] proposes local spatialtemporal graphs that focus on 1-hop spatial neighbors. In this paper, the CGT model employs counterfactuals to efficiently improve TFP interpretability.…”

Section: Related Work a Traffic Flow Predictionmentioning

confidence: 99%

An Urban Traffic Knowledge Graph-Driven Spatial-Temporal Graph Convolutional Network for Traffic Flow Prediction

Yang

2022

Proceedings of the 11th International Joint Conference on Knowledge Graphs

View full text Add to dashboard Cite

Traffic flow prediction (TFP) is a fundamental problem of the Intelligent Transportation System (ITS), as it models the latent spatial-temporal dependency of traffic flow for potential congestion prediction. Recent graph-based models with multiple kinds of attention mechanisms have achieved promising performance. However, existing methods for traffic flow prediction tend to inherit the bias pattern from the dataset and lack interpretability. To this end, we propose a Counterfactual Graph Transformer (CGT) model with an instance-level explainer (e.g., finding the important subgraphs) specifically designed for TFP. We design a perturbation mask generator over input sensor features at the time dimension and the graph structure on the graph transformer module to obtain spatial and temporal counterfactual explanations. By searching the optimal perturbation masks on the input data feature and graph structures, we can obtain the concise and dominant data or graph edge links for the subsequent TFP task. After re-training the utilized graph transformer model after counterfactual perturbation, we can obtain improved and interpretable traffic flow prediction. Extensive results on three real-world public datasets show that CGT can produce reliable explanations and is promising for traffic flow prediction.

show abstract