Graph Neural Network for Traffic Forecasting: A Survey

Weiwei JiangJiayun Luo

article2021Expert systems with applications1,322 citations

Systematizes graph neural network architectures for road, rail, and ride-hailing traffic forecasting, providing a structured taxonomy alongside open-source datasets and benchmarks to guide spatio-temporal transportation research.

Listen

Rapid urbanization has placed intense strain on urban transportation systems, leading to severe traffic congestion, increased vehicle emissions, and lost economic productivity. Effective early intervention through intelligent transportation systems relies heavily on accurate, timely traffic forecasting. However, capturing complex traffic dynamics has historically challenged linear statistical models and standard grid-based deep learning methods, as physical transit networks operate across non-Euclidean, interconnected structures.

The article systematically evaluates the emergence of graph neural networks in traffic forecasting to determine how graph-based deep learning improves spatial and temporal prediction accuracy across modern transit networks.

To conduct this assessment, the authors reviewed 212 academic studies and preprints published between 2018 and 2020. The evaluation categorizes forecasting targets into road-level, region-level, and station-level problems across key domains including traffic flow, speed, travel time, passenger volume, and ride-hailing demand. In addition, the analysis catalogs open-source datasets, software frameworks, and model implementations to evaluate empirical performance across benchmark scenarios.

The article establishes several key findings. First, graph neural networks consistently achieve state-of-the-art predictive performance, systematically outperforming traditional statistical baselines and earlier deep learning approaches. Second, graph convolutional networks and spatial diffusion networks represent the most prevalent architectures for mapping physical road connectivity and directional traffic propagation. Third, dynamic and adaptive graph frameworks demonstrate superior predictive accuracy over rigid, static graph structures by automatically inferring evolving spatial dependencies from live traffic data. Finally, integrating graph convolutions with sequence architectures, such as one-dimensional convolutional networks and attention mechanisms, yields higher computational efficiency and faster training compared to recurrent network designs.

These findings indicate that adopting graph-based predictive architectures can significantly enhance urban traffic management, streamline municipal transit planning, and optimize fleet dispatching for commercial transit platforms. Improved prediction accuracy translates to lower emissions, reduced congestion, and more reliable transit schedules. However, high-performing models often yield narrow performance margins over simpler statistical models while incurring significantly higher computational and deployment costs. Furthermore, the inherent black-box nature of deep neural networks poses interpretability challenges for municipal decision-makers.

Organizations and public agencies should take concrete steps to harness these technologies. Stakeholders should prioritize hybrid architectures that combine graph neural networks with attention mechanisms or data-adaptive modules, especially when modeling complex urban networks. Implementing graph partitioning techniques is essential when scaling these models across city-wide sensor grids. Before broader deployment, organizations must establish standardized benchmarking protocols and centralized data repositories that track evolving infrastructure conditions over long-term observation windows.

Decision-makers must interpret current findings with caution due to persistent limitations in the existing literature. Most surveyed models were trained on small datasets spanning less than a single year and evaluated under controlled, high-quality data conditions. This leaves the models sensitive to missing sensor readings, real-world traffic anomalies, and sudden infrastructure changes. Additional validation via live field pilots is required before deploying these architectures in mission-critical transportation systems.

Nanyang Technological University Tsinghua University deep learning forecasting problems graph neural networks graph neural network graph-based traffic forecasting traffic flow forecasting

arXiv: 2101.11174 jwwthu/GNN4Traffic

Paper: Diffusion Convolutional Recurrent Neural Network: Data-Driven Traffic Forecasting, Yaguang Li et al. (2017). This seminal paper introduces DCRNN, establishing the foundation of traffic forecasting with graph neural networks by integrating diffusion convolutions with recurrent architectures.
Paper: Spatio-temporal Graph Convolutional Neural Network: A Deep Learning Framework for Traffic Forecasting, Bing Yu et al. (2017). This work introduces STGCN, establishing the core framework of combining spectral graph convolutions with temporal convolutions for spatial-temporal traffic modeling.
Paper: Graph WaveNet for Deep Spatial-Temporal Graph Modeling, Zonghan Wu et al. (2019). This paper establishes Graph WaveNet, introducing self-adaptive adjacency matrices and dilated causal convolutions that became standard components in spatial-temporal graph modeling.
Paper: Attention Based Spatial-Temporal Graph Convolutional Networks for Traffic Flow Forecasting, Shengnan Guo et al. (2019). This work formulates the ASTGCN framework, demonstrating how spatial-temporal attention mechanisms can be integrated with graph convolutions for multi-scale traffic flow forecasting.
Paper: GMAN: A Graph Multi-Attention Network for Traffic Prediction, Chuanpan Zheng et al. (2019). This paper presents GMAN, establishing key graph multi-attention mechanisms and encoder-decoder designs for long-term spatial-temporal traffic prediction.
Paper: Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting, Lei Bai et al. (2020). This study introduces AGCRN, a primary milestone reviewed in the survey for adaptive, data-driven graph learning and node-specific traffic pattern extraction.
Paper: Spatial-Temporal Synchronous Graph Convolutional Networks: A New Framework for Spatial-Temporal Network Data Forecasting, Chao Song et al. (2020). This paper establishes STSGCN, introducing localized synchronous spatial-temporal graph modeling techniques covered in the survey.
Paper: T-GCN: A Temporal Graph Convolutional Network for Traffic Prediction, Ling Zhao et al. (2018). This work introduces T-GCN, pioneering the combination of graph convolutional networks and gated recurrent units for traffic speed forecasting on road networks.
Paper: Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networks, Zonghan Wu et al. (2020). This paper introduces MTGNN, providing the broader methodology for multivariate time series forecasting with graph neural networks without predefined structures.
Paper: A Comprehensive Survey on Graph Neural Networks, Zonghan Wu et al. (2019). This comprehensive survey provides the foundational taxonomy and theoretical background of spatial-temporal and convolutional graph neural networks reviewed in the traffic domain.

Paper: Graph Neural Network-Based Anomaly Detection in Multivariate Time Series, Ailin Deng et al. (2021). Extends spatial-temporal graph modeling principles from standard traffic forecasting to unsupervised relationship learning and anomaly detection in multivariate time series.
Paper: TimesNet: Temporal 2D-Variation Modeling for General Time Series Analysis, Haixu Wu et al. (2023). Presents a modern 2D-variation temporal backbone for general time series analysis that offers a compelling architectural alternative to traditional spatial-temporal graph networks on traffic benchmarks.
Paper: iTransformer: Inverted Transformers Are Effective for Time Series Forecasting, Yong Liu et al. (2023). Introduces an inverted Transformer architecture that challenges and extends standard spatial-temporal graph approaches for multivariate time series forecasting across transportation benchmarks.

Cover for Graph Neural Network for Traffic Forecasting: A Survey

Abstract

Traffic forecasting is important for the success of intelligent transportation systems. Deep learning models, including convolution neural networks and recurrent neural networks, have been extensively applied in traffic forecasting problems to model spatial and temporal dependencies. In recent years, to model the graph structures in transportation systems as well as contextual information, graph neural networks have been introduced and have achieved state-of-the-art performance in a series of traffic forecasting problems. In this survey, we review the rapidly growing body of research using different graph neural networks, e.g. graph convolutional and graph attention networks, in various traffic forecasting problems, e.g. road traffic flow and speed forecasting, passenger flow forecasting in urban rail transit systems, and demand forecasting in ride-hailing platforms. We also present a comprehensive list of open data and source resources for each problem and identify future research directions. To the best of our knowledge, this paper is the first comprehensive survey that explores the application of graph neural networks for traffic forecasting problems. We have also created a public GitHub repository where the latest papers, open data, and source resources will be updated.

1 Introduction
2 Related Research Surveys
3 Problems
3.1 Traffic Flow
3.2 Traffic Speed
3.3 Traffic Demand
3.4 Other Problems
4 Graphs and Graph Neural Networks
4.1 Traffic Graphs
4.1.1 Graph Construction
4.1.2 Adjacency Matrix Construction
4.2 Graph Neural Networks
5 Open Data and Source Codes
5.1 Open Data
5.1.1 Traffic Sensor Data
5.1.2 Taxi Data
5.1.3 Ride-hailing Data
5.1.4 Bike Data
5.1.5 Subway Data
5.2 Open Source Codes
5.3 State-of-the-art Performance
6 Challenges and Future Directions
6.1 Challenges
6.1.1 Heterogeneous Data
6.1.2 Multi-task Performance
6.1.3 Practical Implementation
6.1.4 Model Interpretation
6.2 Future Directions
6.2.1 Centralized Data Repository
6.2.2 Traffic Graph Design
6.2.3 Combination with Other Techniques
6.2.4 Applications in Real-World ITS Systems
7 Conclusion
References

Knowls

Knowl 1 — Formal Definition of Traffic Graph and Graph-Based Traffic Forecasting
definition
A traffic graph with node features is formally defined as a graph $G = (V, E, A)$ , where:
- $V$ is a set of $N$ vertices (nodes) representing physical or virtual traffic entities, such as road sensors, road segments, intersections, transit stations, or geographic regions.
- $E$ is the set of edges connecting pairs of vertices in $V$ .
- $A \in \mathbb{R}^{N \times N}$ is the adjacency matrix, where each element $a_{ij}$ denotes the connection weight between node $v_i$ and node $v_j$ . If $A$ is symmetric, $G$ is an undirected graph; if $A$ is asymmetric, $G$ is a directed graph.
At any discrete time step $t$ , the traffic states across all $N$ nodes are represented by the node feature matrix $\chi_t \in \mathbb{R}^{N \times d}$ , where $d$ is the feature dimension (e.g., traffic flow volume, average vehicle speed, occupancy, or demand).

Given a sequence of historical observations over a time window of length $T$ , denoted $\chi = \{\chi_1, \chi_2, \dots, \chi_T\}$ , and optional external context variables $\varepsilon$ (such as weather conditions and calendar events), the graph-based traffic forecasting problem is to learn a mapping function $f$ that predicts future traffic states $y$ :

$y = f(\chi, \varepsilon; G)$

In single-step forecasting, $y = \chi_{T+1}$ is the traffic state at the immediately subsequent time step. In multi-step forecasting, $y = \{\chi_{T+1}, \chi_{T+2}, \dots, \chi_{T+H}\}$ represents the traffic states across a future horizon of $H$ time steps.
Knowl 2 — Classification of Traffic Forecasting Problems
definition
Traffic forecasting problems addressed by graph neural networks are categorized across two dimensions: the predicted traffic state variable and the spatial granularity level.
1. By Traffic State Variable:
  - Traffic Flow: The number of vehicles or passengers traversing a spatial unit in a given interval (e.g., road traffic flow, intersection throughput, transit passenger flow, origin-destination (OD) flows).
  - Traffic Speed: The average vehicle speed across a spatial unit, along with closely correlated measures including travel time, estimated time of arrival, and congestion index.
  - Traffic Demand: Unfulfilled or fulfilled trip requests in shared mobility systems (e.g., ride-hailing demand, taxi demand, bike-sharing demand, shared vehicle demand).
  - Other Traffic Metrics: Secondary or anomalous transport dynamics, including traffic accident frequency, traffic anomalies, parking space availability, vehicle emissions, railway delays, and lane occupancy.
2. By Spatial Granularity Level:
  - Road-Level: Formulated directly on the road network topology, where graph nodes represent road sensors, road segments, road intersections, or individual lanes.
  - Region-Level: Formulated over geographic sub-areas, partitioned either into regular geometric grids or irregular administrative/road-bounded polygons, or pairs of regions representing OD flows.
  - Station-Level: Formulated on discrete physical infrastructure stops, including subway stations, bus stops, railway stations, bike docking hubs, or parking facilities.
Knowl 3 — Taxonomy of Adjacency Matrix Construction Methods in Traffic Graphs
model/method
In graph-based traffic modeling, the adjacency matrix $A \in \mathbb{R}^{N \times N}$ defines how spatial dependencies propagate between $N$ nodes. Four primary categories of adjacency matrices are used:
1. Road-Based Matrices:
  - Connection Matrix: Binary indicator where $a_{ij} = 1$ if node $i$ and node $j$ share a direct physical link in the road or transit network, and $a_{ij} = 0$ otherwise.
  - Transportation Connectivity Matrix: $a_{ij} = 1$ if node $i$ and node $j$ are directly connected via high-speed transit (e.g., subway, motorway) or reachable within a fixed threshold time (e.g., within 5 minutes travel time), and $0$ otherwise.
  - Direction Matrix: $a_{ij}$ encodes the geometric angle or directional orientation between road segments.
2. Distance-Based Matrices:
  - Neighbor Matrix: $a_{ij} = 1$ (or $1/4$ for regular grids) if regions $i$ and $j$ share a geographic boundary, and $0$ otherwise.
  - Distance Decay Matrix: Continuous weights $a_{ij} = g(d_{ij})$ defined as an inverse or Gaussian kernel function of physical distance $d_{ij}$ (e.g., network shortest path length, Euclidean distance, or random walk with restart proximity score).
3. Similarity-Based Matrices:
  - Traffic Pattern Similarity Matrix: $a_{ij}$ equals the statistical correlation coefficient (e.g., Pearson correlation) between the historical time series of nodes $i$ and $j$ .
  - Functional Similarity Matrix: $a_{ij}$ measures the correlation of point-of-interest (POI) distributions or land-use categories between regions $i$ and $j$ .
4. Dynamic Matrices: $A$ is parameterized and dynamically learned from data end-to-end via attention mechanisms or node embeddings, removing the dependence on a pre-defined static graph topology.
Knowl 4 — Spectral Graph Convolution and Renormalized Layer Formulation
equation
For an undirected graph $G = (V, E, A)$ with degree matrix $D \in \mathbb{R}^{N \times N}$ whose diagonal elements are $D_{ii} = \sum_j A_{ij}$ , the unnormalized graph Laplacian is $L = D - A$ , and the normalized graph Laplacian is:

$\tilde{L} = I_N - D^{-\frac{1}{2}} A D^{-\frac{1}{2}}$

where $I_N$ is the $N \times N$ identity matrix.

In the first-order Chebyshev polynomial approximation developed by Kipf and Welling for Graph Convolutional Networks (GCN), the graph convolution operation $*_G$ on node feature matrix $X \in \mathbb{R}^{N \times d}$ is expressed as:

$X *_G = W \left( I_N + D^{-\frac{1}{2}} A D^{-\frac{1}{2}} \right) X$

where $W \in \mathbb{R}^{d \times d'}$ is a learnable parameter weight matrix.

To prevent numerical instability and vanishing or exploding gradients when stacking multiple layers, a renormalization trick is applied:

$X *_G = W \left( \tilde{D}^{-\frac{1}{2}} \tilde{A} \tilde{D}^{-\frac{1}{2}} \right) X$

where $\tilde{A} = A + I_N$ and $\tilde{D}$ is the diagonal degree matrix of $\tilde{A}$ , with $\tilde{D}_{ii} = \sum_j \tilde{A}_{ij}$ .
Knowl 5 — Diffusion Graph Convolution for Directed Traffic Networks
model/method
To handle directed and asymmetric traffic dynamics (such as upstream versus downstream propagation in road networks), diffusion graph convolution models spatial dependency as a bidirectional diffusion process with transition probabilities over directed graphs.

Given a node feature matrix $X \in \mathbb{R}^{N \times d}$ , an adjacency matrix $A \in \mathbb{R}^{N \times N}$ , diagonal out-degree matrix $D_O$ ( $D_{O, ii} = \sum_j A_{ij}$ ), and diagonal in-degree matrix $D_I$ ( $D_{I, ii} = \sum_j A_{ji}$ ), the diffusion convolution operation $*_{DC}$ over $K$ diffusion steps is defined as:

$X *_{DC} = \sum_{k=0}^{K-1} \left( \theta_{k,1} (D_O^{-1} A)^k + \theta_{k,2} (D_I^{-1} A^T)^k \right) X$

where:
- $D_O^{-1} A$ represents the transition probability matrix for the forward (downstream) diffusion process.
- $D_I^{-1} A^T$ represents the transition probability matrix for the reverse (upstream) diffusion process.
- $\theta_{k,1}, \theta_{k,2} \in \mathbb{R}^{d \times d'}$ are learnable parameter weight matrices for diffusion step $k$ .
- $K$ is a hyperparameter specifying the maximum diffusion step horizon.
Knowl 6 — Spatial Message Passing and Graph Attention Operations for Traffic Modeling
equation
Spatial-based graph neural networks propagate information directly across node neighborhoods. Two principal spatial formulations utilized in traffic forecasting are:
1. Message Passing Neural Network (MPNN): For a node $v_i \in V$ with neighbor set $\mathcal{N}(v_i)$ at iteration $t$ , the message aggregation phase and node readout phase are defined as:
  
  $m_{v_i}^{(t)} = \sum_{v_j \in \mathcal{N}(v_i)} \mathcal{M}^{(t)}\left( X_i^{(t-1)}, X_j^{(t-1)}, e_{ij} \right)$
  
  $X_i^{(t)} = \mathcal{U}^{(t)}\left( X_i^{(t-1)}, m_{v_i}^{(t)} \right)$
  
  where $X_i^{(t-1)}$ is the hidden representation of node $v_i$ at iteration $t-1$ , $e_{ij}$ is the feature vector of edge $(v_i, v_j)$ , $\mathcal{M}^{(t)}(\cdot)$ is the message aggregation function, and $\mathcal{U}^{(t)}(\cdot)$ is the node update function.
2. Graph Attention Network (GAT): GAT dynamically computes adaptive attention coefficients between connected nodes across $K$ independent attention heads:
  
  $X_i^{(t)} = \Vert_{k=1}^K \sigma \left( \sum_{v_j \in \mathcal{N}(v_i)} \alpha^k\left( X_i^{(t-1)}, X_j^{(t-1)} \right) W^{(t-1)} X_j^{(t-1)} \right)$
  
  where $\Vert$ denotes vector concatenation, $\sigma(\cdot)$ is an activation function, $W^{(t-1)}$ is a shared parameter matrix, and $\alpha^k(X_i, X_j)$ denotes the normalized scalar attention weight assigned to neighbor $v_j$ by node $v_i$ in the $k$ -th attention head.
Knowl 7 — Taxonomy of Temporal Modeling in Spatiotemporal Graph Neural Networks
model/method
Spatiotemporal Graph Neural Networks (ST-GNNs) combine spatial graph convolutions with temporal sequence modeling modules to simultaneously capture spatial and temporal correlations. Temporal modeling architectures in traffic forecasting are categorized into four paradigms:
1. RNN-Based ST-GNNs: Integrate graph convolution operations directly into Recurrent Neural Network units (e.g., replacing standard linear transformations inside LSTM or GRU cells with GCN or diffusion convolution operators, such as in DCRNN and GC-LSTM). They process sequential dependencies recurrently but have higher computational latency and are susceptible to gradient vanishing/exploding over long prediction horizons.
2. CNN-Based ST-GNNs: Employ 1D temporal convolutions, gated linear units (GLUs), or causal dilated convolutions (such as WaveNet blocks in Graph WaveNet or 1D-CNN layers in STGCN) interleaved with spatial graph convolution blocks. This architecture allows parallel sequence computation and stable training.
3. Attention-Based / Transformer ST-GNNs: Apply multi-head self-attention mechanisms along the temporal dimension (e.g., GMAN, Traffic Transformer, STFGNN). This provides direct access to historical time steps without distance-based signal attenuation.
4. FNN-Based ST-GNNs: Utilize standard Feedforward Neural Networks (multilayer perceptrons) to linearly map temporal feature dimensions, providing a lightweight approach for short-range temporal modeling.

Knowl 8 — Empirical Benchmark Performance Comparison Across Traffic Datasets

data/table

The table reports the performance of representative baseline and advanced spatiotemporal GNN models across standard traffic speed, flow, and demand benchmark datasets. Evaluations are conducted at a 60-minute prediction horizon (unless noted as 30 minutes) using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE).

Dataset	Model	RMSE	MAE	MAPE (%)
METR-LA	DCRNN	7.59	3.60	10.50
	ST-UNet	7.40	3.55	10.00
	Graph WaveNet	7.37	3.53	10.00
	SLCNN	7.20	3.30	9.70
	Traffic Transformer	6.68	3.28	9.08
	STFGNN	6.40	3.18	8.81
PeMS-BAY	DCRNN	4.74	2.07	4.90
	SLCNN	4.53	2.03	4.80
	Graph WaveNet	4.52	1.95	4.63
	GMAN	4.32	1.86	4.31
	Traffic Transformer	4.36	1.77	4.29
	STFGNN	3.74	1.66	3.77
PeMSD4	Graph WaveNet	39.70	25.45	17.29
	STGCN	34.89	21.16	13.83
	DCRNN	33.44	21.22	14.17
	AGCRN	32.26	19.83	12.97
	STFGNN	31.88	19.83	13.02
PeMSD8	Graph WaveNet	31.05	19.13	12.68
	STGCN	27.09	17.50	11.29
	DCRNN	26.36	16.82	10.92
	STFGNN	26.22	16.64	10.60
	AGCRN	25.22	15.95	10.09
TaxiSZ	Graph WaveNet	4.76	3.38	–
	DCRNN	4.64	3.31	–
	T-GCN	4.13	2.79	–
	STGCN	4.13	2.76	–
	AST-GCN	4.10	2.77	–
	A3T-GCN	3.97	2.74	–
TaxiNYC (30 min)	STGCN	22.65	18.46	–
	DCRNN	14.79	8.43	–
	Graph WaveNet	13.07	8.10	–
	CCRNN	9.56	5.50	–
BikeNYC (30 min)	STGCN	3.60	2.76	–
	Graph WaveNet	3.29	1.99	–
	DCRNN	3.21	1.90	–
	CCRNN	2.84	1.74	–

The empirical data demonstrate that advanced models integrating dynamic spatial-temporal graphs, attention mechanisms, or data-adaptive graph learning (such as STFGNN, AGCRN, GMAN, and CCRNN) consistently outperform earlier static spectral or diffusion baselines (STGCN, DCRNN) across traffic speed, flow, and demand forecasting domains.

Knowl 9 — Key Challenges in GNN-Based Traffic Forecasting
limitation
The application of graph neural networks to intelligent transportation systems encounters four primary technical challenges:
1. Heterogeneous Data Quality and Outdated Graph Topologies: Real-world graph structures extracted from mapping platforms (e.g., OpenStreetMap) are frequently out of date or misaligned with physical networks. Physical sensors are susceptible to hardware failures, causing missing and noisy data. Furthermore, anomaly data (e.g., incidents, extreme congestion, holiday surges) are sparse, limiting model generalization during irregular events.
2. Multi-Task Learning Across Incompatible Graph Structures: Real-world traffic management requires simultaneous prediction across multiple modalities (e.g., road network traffic flow alongside subway station passenger volume). Because distinct tasks depend on fundamentally different graph topologies (e.g., continuous road links vs. bipartite OD lines vs. discrete station networks), existing single-graph GNN models cannot be directly applied.
3. Computational Scalability on Large Road Networks: Full-scale highway or citywide road networks contain tens of thousands of nodes and edges. Due to the high memory and compute requirements of multi-hop graph convolutions, models are typically restricted to localized subgraphs, limiting network-wide deployment.
4. Model Interpretability: GNNs function primarily as black boxes. Post-hoc explainability techniques for interpreting which spatial propagation paths or temporal windows drive predictions remain underdeveloped for traffic forecasting.
Knowl 10 — Methodological Future Directions for Traffic Graph Neural Networks
model/method
To address current limitations in traffic forecasting with GNNs, five major methodological directions are identified:
1. Standardized Centralized Data Repositories: Establishing unified benchmark platforms that provide standardized data schemas, graph representations (e.g., standardized GIS and matrix structures), automated evaluation protocols, and long-duration records (>1 year) across non-highway modalities.
2. Transportation Knowledge Graphs: Designing relational knowledge graphs that incorporate semantic domain knowledge—such as land-use classifications, POI distributions, traffic regulations, and event schedules—into graph convolution and message passing layers.
3. Transfer Learning and Meta-Learning: Developing transfer learning frameworks (e.g., transferring pre-trained models from data-dense to data-sparse highway regions) and meta-learning architectures (e.g., generating neural network parameter weights dynamically from graph attributes) to generalize across changing infrastructure and newly built transit stations.
4. Automated Machine Learning (AutoML) and Generative Data Augmentation: Applying AutoML algorithms (e.g., reinforcement learning-based neural architecture search) to automate GNN hyperparameter optimization, and utilizing Generative Adversarial Networks (GANs) to synthesize realistic traffic distributions for data augmentation.
5. Bayesian Graph Networks and Uncertainty Quantification: Combining GNNs with Bayesian inference or quantile regression to produce probabilistic traffic forecasts with confidence intervals, which are critical for risk assessment and traffic anomaly intervention.

Coverage note — None omitted; all core survey taxonomies, mathematical formulations, benchmark performance comparisons, challenges, and future research directions contributed by the authors are fully captured.

References

1.Agafonov, A. (2020). Traffic flow prediction using graph convolution neural networks. In 2020 10th International Conference on Information Science and Technology (ICIST) (pp. 91–95). IEEE.
2.Arjovsky, M., Chintala, S., & Bottou, L. (2017). Wasserstein gan. arXiv preprint arXiv:1701.07875 , .
3.Atwood, J., & Towsley, D. (2016). Diffusion-convolutional neural networks. In NIPS.
4.Bai, J., Zhu, J., Song, Y., Zhao, L., Hou, Z., Du, R., & Li, H. (2021). A3t-gcn: attention temporal graph convolutional network for traffic forecasting. ISPRS International Journal of Geo-Information, 10 , 485.
5.Bai, L., Yao, L., Kanhere, S. S., Wang, X., Liu, W., & Yang, Z. (2019a). Spatio-temporal graph convolutional and recurrent networks for citywide passenger demand prediction. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management (pp. 2293–2296).
6.Bai, L., Yao, L., Kanhere, S. S., Wang, X., & Sheng, Q. Z. (2019b). Stg2seq: spatial-temporal graph to sequence model for multi-step passenger demand forecasting. In Proceedings of the 28th International Joint Conference on Artificial Intelligence (pp. 1981–1987). AAAI Press.
7.Bai, L., Yao, L., Li, C., Wang, X., & Wang, C. (2020). Adaptive graph convolutional recurrent network for traffic forecasting. In Advances in Neural Information Processing Systems.
8.Baldassarre, F., & Azizpour, H. (2019). Explainability techniques for graph convolutional networks. In International Conference on Machine Learning (ICML) Workshops, 2019 Workshop on Learning and Reasoning with Graph-Structured Representations.
9.Barredo-Arrieta, A., Laña, I., & Del Ser, J. (2019). What lies beneath: A note on the explainability of black-box machine learning models for road traffic forecasting. In 2019 IEEE Intelligent Transportation Systems Conference (ITSC) (pp. 2232–2237). IEEE.
10.Bing, H., Zhifeng, X., Yangjie, X., Jinxing, H., & Zhanwu, M. (2020). Integrating semantic zoning information with the prediction of road link speed based on taxi gps data. Complexity, 2020 .
11.Bogaerts, T., Masegosa, A. D., Angarita-Zapata, J. S., Onieva, E., & Hellinckx, P. (2020). A graph cnn-lstm neural network for short and long-term traffic forecasting based on trajectory data. Transportation Research Part C: Emerging Technologies, 112 , 62–77.
12.Boukerche, A., Tao, Y., & Sun, P. (2020). Artificial intelligence-based vehicular traffic flow prediction methods for supporting intelligent transportation systems. Computer Networks, 182 , 107484.
13.Boukerche, A., & Wang, J. (2020a). Machine learning-based traffic prediction models for intelligent transportation systems. Computer Networks, 181 , 107530.
14.Boukerche, A., & Wang, J. (2020b). A performance modeling and analysis of a novel vehicular traffic flow prediction system using a hybrid machine learning-based model. Ad Hoc Networks, .
15.Bruna, J., Zaremba, W., Szlam, A., & LeCun, Y. (2014). Spectral networks and deep locally connected networks on graphs. In 2nd International Conference on Learning Representations, ICLR 2014 .
16.Cai, L., Janowicz, K., Mai, G., Yan, B., & Zhu, R. (2020). Traffic transformer: Capturing the continuity and periodicity of time series for traffic forecasting. Transactions in GIS, .
17.Cao, D., Wang, Y., Duan, J., Zhang, C., Zhu, X., Huang, C., Tong, Y., Xu, B., Bai, J., Tong, J. et al. (2020). Spectral temporal graph neural network for multivariate time-series forecasting. Advances in Neural Information Processing Systems, 33 .
18.Chai, D., Wang, L., & Yang, Q. (2018). Bike flow prediction with multi-graph convolutional networks. In Proceedings of the 26th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (pp. 397–400).
19.Chen, C., Li, K., Teo, S. G., Zou, X., Wang, K., Wang, J., & Zeng, Z. (2019). Gated residual recurrent graph neural networks for traffic prediction. In Proceedings of the AAAI Conference on Artificial Intelligence (pp. 485–492). volume 33.
20.Chen, F., Chen, Z., Biswas, S., Lei, S., Ramakrishnan, N., & Lu, C.-T. (2020a). Graph convolutional networks with kalman filtering for traffic prediction. In Proceedings of the 28th International Conference on Advances in Geographic Information Systems (pp. 135–138).
21.Chen, H., Rossi, R. A., Mahadik, K., & Eldardiry, H. (2020b). A context integrated relational spatio-temporal model for demand and supply forecasting. arXiv preprint arXiv:2009.12469 , .
22.Chen, J., Liao, S., Hou, J., Wang, K., & Wen, J. (2020c). Gst-gcn: A geographic-semantic-temporal graph convolutional network for context-aware traffic flow prediction on graph sequences. In 2020 IEEE International Conference on Systems, Man, and Cybernetics (SMC) (pp. 1604–1609). IEEE.
23.Chen, K., Chen, F., Lai, B., Jin, Z., Liu, Y., Li, K., Wei, L., Wang, P., Tang, Y., Huang, J. et al. (2020d). Dynamic spatio-temporal graph-based cnns for traffic flow prediction. IEEE Access, 8 , 185136–185145.
24.Chen, L., Han, K., Yin, Q., & Cao, Z. (2020e). Gdcrn: Global diffusion convolutional residual network for traffic flow prediction. In International Conference on Knowledge Science, Engineering and Management (pp. 438–449). Springer.
25.Chen, W., Chen, L., Xie, Y., Cao, W., Gao, Y., & Feng, X. (2020f). Multi-range attentive bicomponent graph convolutional network for traffic forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence. volume 34.
26.Chen, X., Zhang, Y., Du, L., Fang, Z., Ren, Y., Bian, K., & Xie, K. (2020g). Tssrgcn: Temporal spectral spatial retrieval graph convolutional network for traffic flow forecasting. In 2020 IEEE International Conference on Data Mining (ICDM). IEEE.
27.Chen, Z., Zhao, B., Wang, Y., Duan, Z., & Zhao, X. (2020h). Multitask learning and gcn-based taxi demand prediction for a traffic road network. Sensors, 20 , 3776.
28.Cirstea, R.-G., Guo, C., & Yang, B. (2019). Graph attention recurrent neural networks for correlated time series forecasting. MileTS19@KDD, .
29.Cui, Z., Henrickson, K., Ke, R., & Wang, Y. (2019). Traffic graph convolutional recurrent neural network: A deep learning framework for network-scale traffic learning and forecasting. IEEE Transactions on Intelligent Transportation Systems, .
30.Cui, Z., Ke, R., Pu, Z., Ma, X., & Wang, Y. (2020a). Learning traffic as a graph: A gated graph wavelet recurrent neural network for network-scale traffic prediction. Transportation Research Part C: Emerging Technologies, 115 , 102620.
31.Cui, Z., Lin, L., Pu, Z., & Wang, Y. (2020b). Graph markov network for traffic forecasting with missing data. Transportation Research Part C: Emerging Technologies, 117 , 102671. URL: http://www.sciencedirect.com/science/article/pii/S0968090X20305866. doi:https://doi.org/10.1016/j.trc.2020.102671.
32.Dai, R., Xu, S., Gu, Q., Ji, C., & Liu, K. (2020). Hybrid spatio-temporal graph convolutional network: Improving traffic prediction with navigation data. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining KDD '20 (p. 3074–3082). New York, NY, USA: Association for Computing Machinery. URL: https://doi.org/10.1145/3394486.3403358. doi:10.1145/3394486.3403358.
33.Davis, N., Raina, G., & Jagannathan, K. (2020). Grids versus graphs: Partitioning space for improved taxi demand-supply forecasts. IEEE Transactions on Intelligent Transportation Systems, .
34.Defferrard, M., Bresson, X., & Vandergheynst, P. (2016). Convolutional neural networks on graphs with fast localized spectral filtering. In Proceedings of the 30th International Conference on Neural Information Processing Systems (pp. 3844–3852).
35.Diao, Z., Wang, X., Zhang, D., Liu, Y., Xie, K., & He, S. (2019). Dynamic spatial-temporal graph convolutional neural networks for traffic forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence (pp. 890–897). volume 33.
36.Du, B., Hu, X., Sun, L., Liu, J., Qiao, Y., & Lv, W. (2020). Traffic demand prediction based on dynamic transition convolutional neural network. IEEE Transactions on Intelligent Transportation Systems, .
37.Fan, X., Xiang, C., Gong, L., He, X., Qu, Y., Amirgholipour, S., Xi, Y., Nanda, P., & He, X. (2020). Deep learning for intelligent traffic sensing and prediction: recent advances and future challenges. CCF Transactions on Pervasive Computing and Interaction, (pp. 1–21).
38.Fang, S., Pan, X., Xiang, S., & Pan, C. (2020a). Meta-msnet: Meta-learning based multi-source data fusion for traffic flow prediction. IEEE Signal Processing Letters, .
39.Fang, S., Zhang, Q., Meng, G., Xiang, S., & Pan, C. (2019). Gstnet: Global spatial-temporal network for traffic flow prediction. In Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, IJCAI-19 (pp. 2286–2293). International Joint Conferences on Artificial Intelligence Organization. URL: https://doi.org/10.24963/ijcai.2019/317. doi:10.24963/ijcai.2019/317.
40.Fang, X., Huang, J., Wang, F., Zeng, L., Liang, H., & Wang, H. (2020b). Constgat: Contextual spatial-temporal graph attention network for travel time estimation at baidu maps. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining KDD '20 (p. 2697–2705). New York, NY, USA: Association for Computing Machinery. URL: https://doi.org/10.1145/3394486.3403320. doi:10.1145/3394486.3403320.
41.Feng, D., Wu, Z., Zhang, J., & Wu, Z. (2020). Dynamic global-local spatial-temporal network for traffic speed prediction. IEEE Access, 8 , 209296–209307.
42.Fu, J., Zhou, W., & Chen, Z. (2020). Bayesian spatio-temporal graph convolutional network for traffic forecasting. arXiv preprint arXiv:2010.07498 , .
43.Fukuda, S., Uchida, H., Fujii, H., & Yamada, T. (2020). Short-term prediction of traffic flow under incident conditions using graph convolutional recurrent neural network and traffic simulation. IET Intelligent Transport Systems, .
44.Ge, L., Li, H., Liu, J., & Zhou, A. (2019a). Temporal graph convolutional networks for traffic speed prediction considering external factors. In 2019 20th IEEE International Conference on Mobile Data Management (MDM) (pp. 234–242). IEEE.
45.Ge, L., Li, H., Liu, J., & Zhou, A. (2019b). Traffic speed prediction with missing data based on tgcn. In 2019 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI) (pp. 522–529). IEEE.
46.Ge, L., Li, S., Wang, Y., Chang, F., & Wu, K. (2020). Global spatial-temporal graph convolutional network for urban traffic speed prediction. Applied Sciences, 10 , 1509.
47.Geng, X., Li, Y., Wang, L., Zhang, L., Yang, Q., Ye, J., & Liu, Y. (2019a). Spatiotemporal multi-graph convolution network for ride-hailing demand forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence (pp. 3656–3663). volume 33.
48.Geng, X., Wu, X., Zhang, L., Yang, Q., Liu, Y., & Ye, J. (2019b). Multi-modal graph interaction for multi-graph convolution network in urban spatiotemporal forecasting. arXiv preprint arXiv:1905.11395 , .
49.George, S., & Santra, A. K. (2020). Traffic prediction using multifaceted techniques: A survey. Wireless Personal Communications, 115 , 1047–1106.
50.Gilmer, J., Schoenholz, S. S., Riley, P. F., Vinyals, O., & Dahl, G. E. (2017). Neural message passing for quantum chemistry. In International Conference on Machine Learning (pp. 1263–1272). PMLR.
51.Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., & Bengio, Y. (2014). Generative adversarial nets. Advances in neural information processing systems, 27 , 2672–2680.
52.Guo, G., & Yuan, W. (2020). Short-term traffic speed forecasting based on graph attention temporal convolutional networks. Neurocomputing, .
53.Guo, J., Song, C., & Wang, H. (2019a). A multi-step traffic speed forecasting model based on graph convolutional lstm. In 2019 Chinese Automation Congress (CAC) (pp. 2466–2471). IEEE.
54.Guo, K., Hu, Y., Qian, Z., Liu, H., Zhang, K., Sun, Y., Gao, J., & Yin, B. (2020a). Optimized graph convolution recurrent neural network for traffic prediction. IEEE Transactions on Intelligent Transportation Systems, .
55.Guo, K., Hu, Y., Qian, Z., Sun, Y., Gao, J., & Yin, B. (2020b). Dynamic graph convolution network for traffic forecasting based on latent network of laplace matrix estimation. IEEE Transactions on Intelligent Transportation Systems, .
56.Guo, K., Hu, Y., Qian, Z. S., Sun, Y., Gao, J., & Yin, B. (2020c). An optimized temporal-spatial gated graph convolution network for traffic forecasting. IEEE Intelligent Transportation Systems Magazine, .
57.Guo, R., Jiang, Z., Huang, J., Tao, J., Wang, C., Li, J., & Chen, L. (2019b). Bikenet: Accurate bike demand prediction using graph neural networks for station rebalancing. In 2019 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI) (pp. 686–693). IEEE.
58.Guo, S., Lin, Y., Feng, N., Song, C., & Wan, H. (2019c). Attention based spatial-temporal graph convolutional networks for traffic flow forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence (pp. 922–929). volume 33.
59.Guopeng, L., Knoop, V. L., & van Lint, H. (2020). Dynamic graph filters networks: A gray-box model for multistep traffic forecasting. In 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) (pp. 1–6). IEEE.
60.Haghighat, A. K., Ravichandra-Mouli, V., Chakraborty, P., Esfandiari, Y., Arabi, S., & Sharma, A. (2020). Applications of deep learning in intelligent transportation systems. Journal of Big Data Analytics in Transportation, 2 , 115–145.
61.Hamilton, W., Ying, Z., & Leskovec, J. (2017). Inductive representation learning on large graphs. In Advances in neural information processing systems (pp. 1024–1034).
62.Han, X., Shen, G., Yang, X., & Kong, X. (2020). Congestion recognition for hybrid urban road systems via digraph convolutional network. Transportation Research Part C: Emerging Technologies, 121 , 102877.
63.Han, Y., Wang, S., Ren, Y., Wang, C., Gao, P., & Chen, G. (2019). Predicting station-level short-term passenger flow in a citywide metro network using spatiotemporal graph convolutional neural networks. ISPRS International Journal of Geo-Information, 8 , 243.
64.Hasanzadeh, A., Liu, X., Duffield, N., & Narayanan, K. R. (2019). Piecewise stationary modeling of random processes over graphs with an application to traffic prediction. In 2019 IEEE International Conference on Big Data (Big Data) (pp. 3779–3788). IEEE.
65.He, S., & Shin, K. G. (2020a). Dynamic flow distribution prediction for urban dockless e-scooter sharing reconfiguration. In Proceedings of The Web Conference 2020 (pp. 133–143).
66.He, S., & Shin, K. G. (2020b). Towards fine-grained flow forecasting: A graph attention approach for bike sharing systems. In Proceedings of The Web Conference 2020 WWW '20 (p. 88–98). New York, NY, USA: Association for Computing Machinery. URL: https://doi.org/10.1145/3366423.3380097. doi:10.1145/3366423.3380097.
67.He, Y., Zhao, Y., Wang, H., & Tsui, K. L. (2020). Gc-lstm: A deep spatiotemporal model for passenger flow forecasting of high-speed rail network. In 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) (pp. 1–6). IEEE.
68.Heglund, J. S., Taleongpong, P., Hu, S., & Tran, H. T. (2020). Railway delay prediction with spatial-temporal graph convolutional networks. In 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) (pp. 1–6). IEEE.
69.Henaff, M., Bruna, J., & LeCun, Y. (2015). Deep convolutional networks on graph-structured data. arXiv preprint arXiv:1506.05163 , .
70.Hong, H., Lin, Y., Yang, X., Li, Z., Fu, K., Wang, Z., Qie, X., & Ye, J. (2020). Heteta: Heterogeneous information network embedding for estimating time of arrival. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining KDD '20 (p. 2444–2454). New York, NY, USA: Association for Computing Machinery. URL: https://doi.org/10.1145/3394486.3403294. doi:10.1145/3394486.3403294.
71.Hu, J., Guo, C., Yang, B., Jensen, C. S., & Chen, L. (2018). Recurrent multi-graph neural networks for travel cost prediction. arXiv preprint arXiv:1811.05157 , .
72.Hu, J., Yang, B., Guo, C., Jensen, C. S., & Xiong, H. (2020). Stochastic origin-destination matrix forecasting using dual-stage graph convolutional, recurrent neural networks. In 2020 IEEE 36th International Conference on Data Engineering (ICDE) (pp. 1417–1428). IEEE.
73.Huang, R., Huang, C., Liu, Y., Dai, G., & Kong, W. (2020a). Lsgcn: Long shortterm traffic prediction with graph convolutional networks. In C. Bessiere (Ed.), Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence, IJCAI-20 (pp. 2355–2361). International Joint Conferences on Artificial Intelligence Organization. URL: https://doi.org/10.24963/ijcai.2020/326. doi:10.24963/ijcai.2020/326 main track.
74.Huang, Y., Zhang, S., Wen, J., & Chen, X. (2020b). Short-term traffic flow prediction based on graph convolutional network embedded lstm. In International Conference on Transportation and Development 2020 (pp. 159–168). American Society of Civil Engineers Reston, VA.
75.James, J. (2019). Online traffic speed estimation for urban road networks with few data: A transfer learning approach. In 2019 IEEE Intelligent Transportation Systems Conference (ITSC) (pp. 4024–4029). IEEE.
76.James, J. (2020). Citywide traffic speed prediction: A geometric deep learning approach. Knowledge-Based Systems, (p. 106592).
77.Jepsen, T. S., Jensen, C. S., & Nielsen, T. D. (2019). Graph convolutional networks for road networks. In Proceedings of the 27th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (pp. 460–463).
78.Jepsen, T. S., Jensen, C. S., & Nielsen, T. D. (2020). Relational fusion networks: Graph convolutional networks for road networks. IEEE Transactions on Intelligent Transportation Systems, .
79.Jia, C., Wu, B., & Zhang, X.-P. (2020). Dynamic spatiotemporal graph neural network with tensor network. arXiv preprint arXiv:2003.08729 , .
80.Jiang, W. (2022). Graph-based deep learning for communication networks: A survey. Computer Communications, 185 , 40–54.
81.Jiang, W., & Zhang, L. (2018). Geospatial data to images: A deep-learning framework for traffic forecasting. Tsinghua Science and Technology, 24 , 52–64.
82.Jin, G., Cui, Y., Zeng, L., Tang, H., Feng, Y., & Huang, J. (2020a). Urban ride-hailing demand prediction with multiple spatio-temporal information fusion network. Transportation Research Part C: Emerging Technologies, 117 , 102665.
83.Jin, G., Xi, Z., Sha, H., Feng, Y., & Huang, J. (2020b). Deep multi-view spatiotemporal virtual graph neural network for significant citywide ride-hailing demand prediction. arXiv preprint arXiv:2007.15189 , .
84.Kang, Z., Xu, H., Hu, J., & Pei, X. (2019). Learning dynamic graph embedding for traffic flow forecasting: A graph self-attentive method. In 2019 IEEE Intelligent Transportation Systems Conference (ITSC) (pp. 2570–2576). IEEE.
85.Ke, J., Feng, S., Zhu, Z., Yang, H., & Ye, J. (2021a). Joint predictions of multi-modal ride-hailing demands: A deep multi-task multi-graph learning-based approach. Transportation Research Part C: Emerging Technologies, 127 , 103063.
86.Ke, J., Qin, X., Yang, H., Zheng, Z., Zhu, Z., & Ye, J. (2021b). Predicting origin-destination ride-sourcing demand with a spatio-temporal encoder-decoder residual multi-graph convolutional network. Transportation Research Part C: Emerging Technologies, 122 , 102858.
87.Kim, S.-S., Chung, M., & Kim, Y.-K. (2020). Urban traffic prediction using congestion diffusion model. In 2020 IEEE International Conference on Consumer Electronics-Asia (ICCE-Asia) (pp. 1–4). IEEE.
88.Kim, T. S., Lee, W. K., & Sohn, S. Y. (2019). Graph convolutional network approach applied to predict hourly bike-sharing demands considering spatial, temporal, and global effects. PLOS ONE, 14 , e0220782.
89.Kipf, T. N., & Welling, M. (2016). Variational graph auto-encoders. arXiv preprint arXiv:1611.07308 , .
90.Kipf, T. N., & Welling, M. (2017). Semi-supervised classification with graph convolutional networks. In International Conference on Learning Representations (ICLR '17).
91.Kong, X., Xing, W., Wei, X., Bao, P., Zhang, J., & Lu, W. (2020). Stgat: Spatial-temporal graph attention networks for traffic flow forecasting. IEEE Access, .
92.Lee, D., Jung, S., Cheon, Y., Kim, D., & You, S. (2019). Demand forecasting from spatiotemporal data with graph networks and temporal-guided embedding. arXiv preprint arXiv:1905.10709 .
93.Lee, K., Eo, M., Jung, E., Yoon, Y., & Rhee, W. (2021). Short-term traffic prediction with deep neural networks: A survey. IEEE Access, 9 , 54739–54756.
94.Lee, K., & Rhee, W. (2019). Graph convolutional modules for traffic forecasting. CoRR, abs/1905.12256 . URL: http://arxiv.org/abs/1905.12256. arXiv:1905.12256.
95.Lee, K., & Rhee, W. (2022). Ddp-gcn: Multi-graph convolutional network for spatiotemporal traffic forecasting. Transportation Research Part C: Emerging Technologies, 134 , 103466.
96.Lewenfus, G., Martins, W. A., Chatzinotas, S., & Ottersten, B. (2020). Joint forecasting and interpolation of time-varying graph signals using deep learning. IEEE Transactions on Signal and Information Processing over Networks , .
97.Li, A., & Axhausen, K. W. (2020). Short-term traffic demand prediction using graph convolutional neural networks. AGILE: GIScience Series, 1 , 1–14.
98.Li, C., Bai, L., Liu, W., Yao, L., & Waller, S. T. (2020a). Knowledge adaption for demand prediction based on multi-task memory neural network. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management (pp. 715–724).
99.Li, J., Peng, H., Liu, L., Xiong, G., Du, B., Ma, H., Wang, L., & Bhuiyan, M. Z. A. (2018a). Graph cnns for urban traffic passenger flows prediction. In 2018 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI) (pp. 29–36). IEEE.
100.Li, L., Yan, J., Yang, X., & Jin, Y. (2019a). Learning interpretable deep state space model for probabilistic time series forecasting. In Proceedings of the 28th International Joint Conference on Artificial Intelligence (pp. 2901–2908).
101.Li, M., & Zhu, Z. (2021). Spatial-temporal fusion graph neural networks for traffic flow forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence (pp. 4189–4196). volume 35.
102.Li, W., Wang, X., Zhang, Y., & Wu, Q. (2020b). Traffic flow prediction over muti-sensor data correlation with graph convolution network. Neurocomputing, .
103.Li, W., Yang, X., Tang, X., & Xia, S. (2020c). Sdcn: Sparsity and diversity driven correlation networks for traffic demand forecasting. In 2020 International Joint Conference on Neural Networks (IJCNN) (pp. 1–8). IEEE.
104.Li, Y., & Moura, J. M. (2020). Forecaster: A graph transformer for forecasting spatial and time-dependent data. In Proceedings of the Twenty-fourth European Conference on Artificial Intelligence.
105.Li, Y., Yu, R., Shahabi, C., & Liu, Y. (2018b). Diffusion convolutional recurrent neural network: Data-driven traffic forecasting. In International Conference on Learning Representations (ICLR '18).
106.Li, Z., Li, L., Peng, Y., & Tao, X. (2020d). A two-stream graph convolutional neural network for dynamic traffic flow forecasting. In 2020 IEEE 32nd International Conference on Tools with Artificial Intelligence (ICTAI) (pp. 355–362). IEEE.
107.Li, Z., Sergin, N. D., Yan, H., Zhang, C., & Tsung, F. (2020e). Tensor completion for weakly-dependent data on graph for metro passenger flow prediction. In Proceedings of the AAAI Conference on Artificial Intelligence. volume 34.
108.Li, Z., Xiong, G., Chen, Y., Lv, Y., Hu, B., Zhu, F., & Wang, F.-Y. (2019b). A hybrid deep learning approach with gcn and lstm for traffic flow prediction. In 2019 IEEE Intelligent Transportation Systems Conference (ITSC) (pp. 1929–1933). IEEE.
109.Li, Z., Xiong, G., Tian, Y., Lv, Y., Chen, Y., Hui, P., & Su, X. (2020f). A multi-stream feature fusion approach for traffic prediction. IEEE Transactions on Intelligent Transportation Systems, .
110.Liao, B., Zhang, J., Wu, C., McIlwraith, D., Chen, T., Yang, S., Guo, Y., & Wu, F. (2018). Deep sequence learning with auxiliary information for traffic prediction. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (pp. 537–546).
111.Lin, L., He, Z., & Peeta, S. (2018). Predicting station-level hourly demand in a large-scale bike-sharing network: A graph convolutional neural network approach. Transportation Research Part C: Emerging Technologies, 97 , 258–276.
112.Liu, J., Ong, G. P., & Chen, X. (2020a). Graphsage-based traffic speed forecasting for segment network with sparse data. IEEE Transactions on Intelligent Transportation Systems, .
113.Liu, L., Chen, J., Wu, H., Zhen, J., Li, G., & Lin, L. (2020b). Physical-virtual collaboration modeling for intra-and inter-station metro ridership prediction. IEEE Transactions on Intelligent Transportation Systems, .
114.Liu, L., Zhou, T., Long, G., Jiang, J., & Zhang, C. (2019). Learning to propagate for graph meta-learning. In Advances in Neural Information Processing Systems (pp. 1039–1050).
115.Liu, R., Zhao, S., Cheng, B., Yang, H., Tang, H., & Yang, F. (2020c). St-mfm: A spatiotemporal multi-modal fusion model for urban anomalies prediction. In Proceedings of the Twenty-fourth European Conference on Artificial Intelligence.
116.Lu, B., Gan, X., Jin, H., Fu, L., & Zhang, H. (2020a). Spatiotemporal adaptive gated graph convolution network for urban traffic flow forecasting. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management (pp. 1025–1034).
117.Lu, M., Zhang, K., Liu, H., & Xiong, N. (2019a). Graph hierarchical convolutional recurrent neural network (ghcrnn) for vehicle condition prediction. arXiv preprint arXiv:1903.06261 , .
118.Lu, Z., Lv, W., Cao, Y., Xie, Z., Peng, H., & Du, B. (2020b). Lstm variants meet graph neural networks for road speed prediction. Neurocomputing, .
119.Lu, Z., Lv, W., Xie, Z., Du, B., & Huang, R. (2019b). Leveraging graph neural network with lstm for traffic speed prediction. In 2019 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI) (pp. 74–81). IEEE.
120.Luca, M., Barlacchi, G., Lepri, B., & Pappalardo, L. (2020). Deep learning for human mobility: a survey on data and models. arXiv preprint arXiv:2012.02825 , .
121.Luo, M., Du, B., Klemmer, K., Zhu, H., Ferhatosmanoglu, H., & Wen, H. (2020). D3p: Data-driven demand prediction for fast expanding electric vehicle sharing systems. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 4 , 1–21.
122.Lv, M., Hong, Z., Chen, L., Chen, T., Zhu, T., & Ji, S. (2020). Temporal multi-graph convolutional network for traffic flow prediction. IEEE Transactions on Intelligent Transportation Systems, .
123.Maas, T., & Bloem, P. (2020). Uncertainty intervals for graph-based spatio-temporal traffic prediction. arXiv preprint arXiv:2012.05207 , .
124.Mallick, T., Balaprakash, P., Rask, E., & Macfarlane, J. (2020). Graph-partitioning-based diffusion convolution recurrent neural network for large-scale traffic forecasting. Transportation Research Record, (p. 0361198120930010).
125.Mallick, T., Balaprakash, P., Rask, E., & Macfarlane, J. (2021). Transfer learning with graph neural networks for short-term highway traffic forecasting. In 2020 25th International Conference on Pattern Recognition (ICPR) (pp. 10367–10374). IEEE.
126.Manibardo, E. L., Laña, I., & Del Ser, J. (2021). Deep learning for road traffic forecasting: Does it make a difference? IEEE Transactions on Intelligent Transportation Systems, .
127.Mena-Oreja, J., & Gozalvez, J. (2020). A comprehensive evaluation of deep learning-based techniques for traffic prediction. IEEE Access, 8 , 91188–91212.
128.Mohanty, S., & Pozdnukhov, A. (2018). Graph cnn+ lstm framework for dynamic macroscopic traffic congestion prediction. In International Workshop on Mining and Learning with Graphs.
129.Mohanty, S., Pozdnukhov, A., & Cassidy, M. (2020). Region-wide congestion prediction and control using deep learning. Transportation Research Part C: Emerging Technologies, 116 , 102624.
130.Opolka, F. L., Solomon, A., Cangea, C., Veličković, P., Liò, P., & Hjelm, R. D. (2019). Spatio-temporal deep graph infomax. In Representation Learning on Graphs and Manifolds, ICLR 2019 Workshop.
131.Oreshkin, B. N., Amini, A., Coyle, L., & Coates, M. (2021). Fc-gaga: Fully connected gated graph architecture for spatio-temporal traffic forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence (pp. 9233–9241). volume 35.
132.Ou, J., Sun, J., Zhu, Y., Jin, H., Liu, Y., Zhang, F., Huang, J., & Wang, X. (2020). Stp-trellisnets: Spatial-temporal parallel trellisnets for metro station passenger flow prediction. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management (pp. 1185–1194).
133.Pan, Z., Liang, Y., Wang, W., Yu, Y., Zheng, Y., & Zhang, J. (2019). Urban traffic prediction from spatio-temporal data using deep meta learning. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (pp. 1720–1730).
134.Pan, Z., Zhang, W., Liang, Y., Zhang, W., Yu, Y., Zhang, J., & Zheng, Y. (2020). Spatio-temporal meta learning for urban traffic prediction. IEEE Transactions on Knowledge and Data Engineering, .
135.Park, C., Lee, C., Bahng, H., Tae, Y., Jin, S., Kim, K., Ko, S., & Choo, J. (2020). St-grat: A novel spatio-temporal graph attention networks for accurately forecasting dynamically changing road speed. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management (pp. 1215–1224).
136.Pavlyuk, D. (2019). Feature selection and extraction in spatiotemporal traffic forecasting: a systematic literature review. European Transport Research Review, 11 , 6.
137.Peng, H., Wang, H., Du, B., Bhuiyan, M. Z. A., Ma, H., Liu, J., Wang, L., Yang, Z., Du, L., Wang, S. et al. (2020). Spatial temporal incidence dynamic graph neural networks for traffic flow forecasting. Information Sciences, 521 , 277–290.
138.Pian, W., & Wu, Y. (2020). Spatial-temporal dynamic graph attention networks for ride-hailing demand prediction. arXiv preprint arXiv:2006.05905 , .
139.Pope, P. E., Kolouri, S., Rostami, M., Martin, C. E., & Hoffmann, H. (2019). Explainability methods for graph convolutional neural networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 10772–10781).
140.Qin, K., Xu, Y., Kang, C., & Kwan, M.-P. (2020a). A graph convolutional network model for evaluating potential congestion spots based on local urban built environments. Transactions in GIS, .
141.Qin, T., Liu, T., Wu, H., Tong, W., & Zhao, S. (2020b). Resgcn: Residual graph convolutional network based free dock prediction in bike sharing system. In 2020 21st IEEE International Conference on Mobile Data Management (MDM) (pp. 210–217). IEEE.
142.Qiu, H., Zheng, Q., Msahli, M., Memmi, G., Qiu, M., & Lu, J. (2020). Topological graph convolutional network-based urban traffic flow and density prediction. IEEE Transactions on Intelligent Transportation Systems, .
143.Qu, Y., Zhu, Y., Zang, T., Xu, Y., & Yu, J. (2020). Modeling local and global flow aggregation for traffic flow forecasting. In International Conference on Web Information Systems Engineering (pp. 414–429). Springer.
144.Ramadan, A., Elbery, A., Zorba, N., & Hassanein, H. S. (2020). Traffic forecasting using temporal line graph convolutional network: Case study. In ICC 2020-2020 IEEE International Conference on Communications (ICC) (pp. 1–6). IEEE.
145.Ren, Y., & Xie, K. (2019). Transfer knowledge between sub-regions for traffic prediction using deep learning method. In International Conference on Intelligent Data Engineering and Automated Learning (pp. 208–219). Springer.
146.Sánchez, C. S., Wieder, A., Sottovia, P., Bortoli, S., Baumbach, J., & Axenie, C. (2020). Gannster: Graph-augmented neural network spatio-temporal reasoner for traffic forecasting. In International Workshop on Advanced Analysis and Learning on Temporal Data (AALTD). Springer.
147.Satorras, V. G., & Estrach, J. B. (2018). Few-shot learning with graph neural networks. In International Conference on Learning Representations.
148.Scarselli, F., Gori, M., Tsoi, A. C., Hagenbuchner, M., & Monfardini, G. (2008). The graph neural network model. IEEE transactions on neural networks, 20 , 61–80.
149.Shao, K., Wang, K., Chen, L., & Zhou, Z. (2020). Estimation of urban travel time with sparse traffic surveillance data. In Proceedings of the 2020 4th High Performance Computing and Cluster Technologies Conference & 2020 3rd International Conference on Big Data and Artificial Intelligence (pp. 218–223).
150.Shen, Y., Jin, C., & Hua, J. (2020). Ttpnet: A neural network for travel time prediction based on tensor decomposition and graph embedding. IEEE Transactions on Knowledge and Data Engineering, .
151.Shi, H., Yao, Q., Guo, Q., Li, Y., Zhang, L., Ye, J., Li, Y., & Liu, Y. (2020). Predicting origin-destination flow via multi-perspective graph convolutional network. In 2020 IEEE 36th International Conference on Data Engineering (ICDE) (pp. 1818–1821). IEEE.
152.Shi, X., & Yeung, D.-Y. (2018). Machine learning for spatiotemporal sequence forecasting: A survey. arXiv preprint arXiv:1808.06865 , .
153.Shin, Y., & Yoon, Y. (2020). Incorporating dynamicity of transportation network with multi-weight traffic graph convolutional network for traffic forecasting. IEEE Transactions on Intelligent Transportation Systems, .
154.Shleifer, S., McCreery, C., & Chitters, V. (2019). Incrementally improving graph wavenet performance on traffic prediction. arXiv preprint arXiv:1912.07390 , .
155.Song, C., Lin, Y., Guo, S., & Wan, H. (2020a). Spatial-temporal synchronous graph convolutional networks: A new framework for spatial-temporal network data forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence (pp. 914–921). volume 34.
156.Song, Q., Ming, R., Hu, J., Niu, H., & Gao, M. (2020b). Graph attention convolutional network: Spatiotemporal modeling for urban traffic prediction. In 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) (pp. 1–6). IEEE.
157.Sun, J., Zhang, J., Li, Q., Yi, X., Liang, Y., & Zheng, Y. (2020). Predicting citywide crowd flows in irregular regions using multi-view graph convolutional networks. IEEE Transactions on Knowledge and Data Engineering, (pp. 1–1).
158.Sun, X., Li, J., Lv, Z., & Dong, C. (2020). Traffic flow prediction model based on spatio-temporal dilated graph convolution. KSII Transactions on Internet & Information Systems, 14 .
159.Sun, Y., Wang, Y., Fu, K., Wang, Z., Zhang, C., & Ye, J. (2021). Constructing geographic and long-term temporal graph for traffic forecasting. In 2020 25th International Conference on Pattern Recognition (ICPR) (pp. 3483–3490). IEEE.
160.Tang, C., Sun, J., & Sun, Y. (2020a). Dynamic spatial-temporal graph attention graph convolutional network for short-term traffic flow forecasting. In 2020 IEEE International Symposium on Circuits and Systems (ISCAS) (pp. 1–5). IEEE.
161.Tang, C., Sun, J., Sun, Y., Peng, M., & Gan, N. (2020b). A general traffic flow prediction approach based on spatial-temporal graph attention. IEEE Access, 8 , 153731–153741.
162.Tedjopurnomo, D. A., Bao, Z., Zheng, B., Choudhury, F., & Qin, A. (2020). A survey on modern deep neural network for traffic prediction: Trends, methods and challenges. IEEE Transactions on Knowledge and Data Engineering, .
163.Tian, K., Guo, J., Ye, K., & Xu, C.-Z. (2020). St-mgat: Spatial-temporal multi-head graph attention networks for traffic forecasting. In 2020 IEEE 32nd International Conference on Tools with Artificial Intelligence (ICTAI) (pp. 714–721). IEEE.
164.Varghese, V., Chikaraishi, M., & Urata, J. (2020). Deep learning in transport studies: A meta-analysis on the prediction accuracy. Journal of Big Data Analytics in Transportation, (pp. 1–22).
165.Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017). Attention is all you need. Advances in neural information processing systems, 30 , 5998–6008.
166.Veličković, P., Cucurull, G., Casanova, A., Romero, A., Liò, P., & Bengio, Y. (2018). Graph attention networks. In International Conference on Learning Representations.
167.Wang, B., Luo, X., Zhang, F., Yuan, B., Bertozzi, A. L., & Brantingham, P. J. (2018a). Graph-based deep modeling and real time forecasting of sparse spatio-temporal data. arXiv preprint arXiv:1804.00684 , .
168.Wang, C., Zhang, K., Wang, H., & Chen, B. (2020a). Auto-stgcn: Autonomous spatial-temporal graph convolutional network search based on reinforcement learning and existing research results. arXiv preprint arXiv:2010.07474 , .
169.Wang, F., Xu, J., Liu, C., Zhou, R., & Zhao, P. (2020b). Mtgcn: A multitask deep learning model for traffic flow prediction. In International Conference on Database Systems for Advanced Applications (pp. 435–451). Springer.
170.Wang, H.-W., Peng, Z.-R., Wang, D., Meng, Y., Wu, T., Sun, W., & Lu, Q.-C. (2020c). Evaluation and prediction of transportation resilience under extreme weather events: A diffusion graph convolutional approach. Transportation Research Part C: Emerging Technologies, 115 , 102619.
171.Wang, J., Kong, L., Huang, Z., & Xiao, J. (2021). Survey of graph neural network. Computer Engineering, 47 , 1–12.
172.Wang, Q., Guo, B., Ouyang, Y., Shu, K., Yu, Z., & Liu, H. (2020d). Spatial community-informed evolving graphs for demand prediction. In Proceedings of The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD 2020).
173.Wang, S., Miao, H., Chen, H., & Huang, Z. (2020e). Multi-task adversarial spatial-temporal networks for crowd flow prediction. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management (pp. 1555–1564).
174.Wang, X., Chen, C., Min, Y., He, J., Yang, B., & Zhang, Y. (2018b). Efficient metropolitan traffic prediction based on graph recurrent neural network. arXiv preprint arXiv:1811.00740 , .
175.Wang, X., Guan, X., Cao, J., Zhang, N., & Wu, H. (2020f). Forecast network-wide traffic states for multiple steps ahead: A deep learning approach considering dynamic non-local spatial correlation and non-stationary temporal dependency. Transportation Research Part C: Emerging Technologies, 119 , 102763. URL: http://www.sciencedirect.com/science/article/pii/S0968090X20306756. doi:https://doi.org/10.1016/j.trc.2020.102763.
176.Wang, X., Ma, Y., Wang, Y., Jin, W., Wang, X., Tang, J., Jia, C., & Yu, J. (2020g). Traffic flow prediction via spatial temporal graph neural network. In Proceedings of The Web Conference 2020 WWW '20 (p. 1082–1092). New York, NY, USA: Association for Computing Machinery. URL: https://doi.org/10.1145/3366423.3380186. doi:10.1145/3366423.3380186.
177.Wang, Y., Xu, D., Peng, P., Xuan, Q., & Zhang, G. (2020h). An urban commuters' od hybrid prediction method based on big gps data. Chaos: An Interdisciplinary Journal of Nonlinear Science, 30 , 093128.
178.Wang, Y., Yin, H., Chen, H., Wo, T., Xu, J., & Zheng, K. (2019). Origin-destination matrix prediction via graph convolution: a new perspective of passenger demand modeling. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (pp. 1227–1235).
179.Wei, C., & Sheng, J. (2020). Spatial-temporal graph attention networks for traffic flow forecasting. In IOP Conference Series: Earth and Environmental Science (p. 012065). IOP Publishing volume 587.
180.Wei, L., Yu, Z., Jin, Z., Xie, L., Huang, J., Cai, D., He, X., & Hua, X.-S. (2019). Dual graph for traffic forecasting. IEEE Access, .
181.Wright, M. A., Ehlers, S. F., & Horowitz, R. (2019). Neural-attention-based deep learning architectures for modeling traffic dynamics on lane graphs. In 2019 IEEE Intelligent Transportation Systems Conference (ITSC) (pp. 3898–3905). IEEE.
182.Wu, M., Zhu, C., & Chen, L. (2020a). Multi-task spatial-temporal graph attention network for taxi demand prediction. In Proceedings of the 2020 5th International Conference on Mathematics and Artificial Intelligence (pp. 224–228).
183.Wu, T., Chen, F., & Wan, Y. (2018a). Graph attention lstm network: A new model for traffic flow forecasting. In 2018 5th International Conference on Information Science and Control Engineering (ICISCE) (pp. 241–245). IEEE.
184.Wu, Y., Tan, H., Qin, L., Ran, B., & Jiang, Z. (2018b). A hybrid deep learning based traffic flow prediction method and its understanding. Transportation Research Part C: Emerging Technologies, 90 , 166–180.
185.Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., & Philip, S. Y. (2020b). A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems, .
186.Wu, Z., Pan, S., Long, G., Jiang, J., Chang, X., & Zhang, C. (2020c). 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 KDD '20 (p. 753–763). New York, NY, USA: Association for Computing Machinery. URL: https://doi.org/10.1145/3394486.3403118. doi:10.1145/3394486.3403118.
187.Wu, Z., Pan, S., Long, G., Jiang, J., & Zhang, C. (2019). Graph wavenet for deep spatial-temporal graph modeling. In Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, IJCAI-19 (pp. 1907–1913). International Joint Conferences on Artificial Intelligence Organization. URL: https://doi.org/10.24963/ijcai.2019/264. doi:10.24963/ijcai.2019/264.
188.Xiao, G., Wang, R., Zhang, C., & Ni, A. (2020). Demand prediction for a public bike sharing program based on spatio-temporal graph convolutional networks. Multimedia Tools and Applications, (pp. 1–19).
189.Xie, P., Li, T., Liu, J., Du, S., Yang, X., & Zhang, J. (2020a). Urban flow prediction from spatiotemporal data using machine learning: A survey. Information Fusion, 59 , 1–12.
190.Xie, Q., Guo, T., Chen, Y., Xiao, Y., Wang, X., & Zhao, B. Y. (2020b). Deep graph convolutional networks for incident-driven traffic speed prediction. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management (pp. 1665–1674).
191.Xie, Y., Xiong, Y., & Zhu, Y. (2020c). Istd-gcn: Iterative spatial-temporal diffusion graph convolutional network for traffic speed forecasting. arXiv preprint arXiv:2008.03970 , .
192.Xie, Y., Xiong, Y., & Zhu, Y. (2020d). Sast-gnn: A self-attention based spatio-temporal graph neural network for traffic prediction. In International Conference on Database Systems for Advanced Applications (pp. 707–714). Springer.
193.Xie, Z., Lv, W., Huang, S., Lu, Z., Du, B., & Huang, R. (2019). Sequential graph neural network for urban road traffic speed prediction. IEEE Access, .
194.Xin, Y., Miao, D., Zhu, M., Jin, C., & Lu, X. (2020). Internet: Multistep traffic forecasting by interacting spatial and temporal features. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management (pp. 3477–3480).
195.Xiong, X., Ozbay, K., Jin, L., & Feng, C. (2020). Dynamic origin–destination matrix prediction with line graph neural networks and kalman filter. Transportation Research Record, (p. 0361198120919399).
196.Xu, D., Dai, H., Wang, Y., Peng, P., Xuan, Q., & Guo, H. (2019). Road traffic state prediction based on a graph embedding recurrent neural network under the scats. Chaos: An Interdisciplinary Journal of Nonlinear Science, 29 , 103125.
197.Xu, D., Wei, C., Peng, P., Xuan, Q., & Guo, H. (2020a). Ge-gan: A novel deep learning framework for road traffic state estimation. Transportation Research Part C: Emerging Technologies, 117 , 102635.
198.Xu, M., Dai, W., Liu, C., Gao, X., Lin, W., Qi, G.-J., & Xiong, H. (2020b). Spatial-temporal transformer networks for traffic flow forecasting. arXiv preprint arXiv:2001.02908 , .
199.Xu, X., Zheng, H., Feng, X., & Chen, Y. (2020c). Traffic flow forecasting with spatial-temporal graph convolutional networks in edge-computing systems. In 2020 International Conference on Wireless Communications and Signal Processing (WCSP) (pp. 251–256). IEEE.
200.Xu, Y., & Li, D. (2019). Incorporating graph attention and recurrent architectures for city-wide taxi demand prediction. ISPRS International Journal of Geo-Information, 8 , 414.
201.Xu, Z., Kang, Y., Cao, Y., & Li, Z. (2020d). Spatiotemporal graph convolution multifusion network for urban vehicle emission prediction. IEEE Transactions on Neural Networks and Learning Systems, .
202.Yang, F., Chen, L., Zhou, F., Gao, Y., & Cao, W. (2020). Relational state-space model for stochastic multi-object systems. In International Conference on Learning Representations.
203.Yang, S., Ma, W., Pi, X., & Qian, S. (2019). A deep learning approach to real-time parking occupancy prediction in transportation networks incorporating multiple spatio-temporal data sources. Transportation Research Part C: Emerging Technologies, 107 , 248–265.
204.Yao, X., Gao, Y., Zhu, D., Manley, E., Wang, J., & Liu, Y. (2020). Spatial origin-destination flow imputation using graph convolutional networks. IEEE Transactions on Intelligent Transportation Systems, .
205.Ye, J., Sun, L., Du, B., Fu, Y., & Xiong, H. (2021). Coupled layer-wise graph convolution for transportation demand prediction. In Proceedings of the AAAI Conference on Artificial Intelligence (pp. 4617–4625). volume 35.
206.Ye, J., Zhao, J., Ye, K., & Xu, C. (2020a). How to build a graph-based deep learning architecture in traffic domain: A survey. IEEE Transactions on Intelligent Transportation Systems, .
207.Ye, J., Zhao, J., Ye, K., & Xu, C. (2020b). Multi-stgcnet: A graph convolution based spatial-temporal framework for subway passenger flow forecasting. In 2020 International Joint Conference on Neural Networks (IJCNN) (pp. 1–8). IEEE.
208.Yeghikyan, G., Opolka, F. L., Nanni, M., Lepri, B., & Liò, P. (2020). Learning mobility flows from urban features with spatial interaction models and neural networks. In 2020 IEEE International Conference on Smart Computing (SMARTCOMP) (pp. 57–64). IEEE.
209.Yin, X., Wu, G., Wei, J., Shen, Y., Qi, H., & Yin, B. (2020). Multi-stage attention spatial-temporal graph networks for traffic prediction. Neurocomputing, .
210.Yin, X., Wu, G., Wei, J., Shen, Y., Qi, H., & Yin, B. (2021). Deep learning on traffic prediction: Methods, analysis and future directions. IEEE Transactions on Intelligent Transportation Systems, .
211.Ying, Z., Bourgeois, D., You, J., Zitnik, M., & Leskovec, J. (2019). Gnnexplainer: Generating explanations for graph neural networks. In Advances in neural information processing systems (pp. 9244–9255).
212.Yoshida, A., Yatsushiro, Y., Hata, N., Higurashi, T., Tateiwa, N., Wakamatsu, T., Tanaka, A., Nagamatsu, K., & Fujisawa, K. (2019). Practical end-to-end repositioning algorithm for managing bike-sharing system. In 2019 IEEE International Conference on Big Data (Big Data) (pp. 1251–1258). IEEE.
213.Yu, B., Lee, Y., & Sohn, K. (2020a). Forecasting road traffic speeds by considering area-wide spatio-temporal dependencies based on a graph convolutional neural network (gcn). Transportation Research Part C: Emerging Technologies, 114 , 189–204.
214.Yu, B., Li, M., Zhang, J., & Zhu, Z. (2019a). 3d graph convolutional networks with temporal graphs: A spatial information free framework for traffic forecasting. arXiv preprint arXiv:1903.00919 , .
215.Yu, B., Yin, H., & 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, IJCAI-18 (pp. 3634–3640). International Joint Conferences on Artificial Intelligence Organization. URL: https://doi.org/10.24963/ijcai.2018/505. doi:10.24963/ijcai.2018/505.
216.Yu, B., Yin, H., & Zhu, Z. (2019b). St-unet: A spatio-temporal u-network for graph-structured time series modeling. arXiv preprint arXiv:1903.05631 , .
217.Yu, J. J. Q., & Gu, J. (2019). Real-time traffic speed estimation with graph convolutional generative autoencoder. IEEE Transactions on Intelligent Transportation Systems, 20 , 3940–3951.
218.Yu, L., Du, B., Hu, X., Sun, L., Han, L., & Lv, W. (2020b). Deep spatio-temporal graph convolutional network for traffic accident prediction. Neurocomputing, .
219.Yuan, J., Zheng, Y., Zhang, C., Xie, W., Xie, X., Sun, G., & Huang, Y. (2010). T-drive: driving directions based on taxi trajectories. In Proceedings of the 18th SIGSPATIAL International conference on advances in geographic information systems (pp. 99–108).
220.Zhang, C., James, J., & Liu, Y. (2019a). Spatial-temporal graph attention networks: A deep learning approach for traffic forecasting. IEEE Access, 7 , 166246–166256.
221.Zhang, H., Liu, J., Tang, Y., & Xiong, G. (2020a). Attention based graph covolution networks for intelligent traffic flow analysis. In 2020 IEEE 16th International Conference on Automation Science and Engineering (CASE) (pp. 558–563). IEEE.
222.Zhang, J., Chen, F., Cui, Z., Guo, Y., & Zhu, Y. (2020b). Deep learning architecture for short-term passenger flow forecasting in urban rail transit. IEEE Transactions on Intelligent Transportation Systems, .
223.Zhang, J., Chen, F., & Guo, Y. (2020c). Multi-graph convolutional network for short-term passenger flow forecasting in urban rail transit. IET Intelligent Transport Systems, .
224.Zhang, J., Shi, X., Xie, J., Ma, H., King, I., & Yeung, D. Y. (2018a). Gaan: Gated attention networks for learning on large and spatiotemporal graphs. In 34th Conference on Uncertainty in Artificial Intelligence 2018, UAI 2018 .
225.Zhang, J., Zheng, Y., & Qi, D. (2017). Deep spatio-temporal residual networks for citywide crowd flows prediction. In Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence (pp. 1655–1661).
226.Zhang, K., He, F., Zhang, Z., Lin, X., & Li, M. (2020d). Graph attention temporal convolutional network for traffic speed forecasting on road networks. Transportmetrica B: Transport Dynamics, (pp. 1–19).
227.Zhang, N., Guan, X., Cao, J., Wang, X., & Wu, H. (2019b). A hybrid traffic speed forecasting approach integrating wavelet transform and motif-based graph convolutional recurrent neural network. arXiv preprint arXiv:1904.06656 , .
228.Zhang, Q., Chang, J., Meng, G., Xiang, S., & Pan, C. (2020e). Spatio-temporal graph structure learning for traffic forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence. volume 34.
229.Zhang, Q., Jin, Q., Chang, J., Xiang, S., & Pan, C. (2018b). Kernel-weighted graph convolutional network: A deep learning approach for traffic forecasting. In 2018 24th International Conference on Pattern Recognition (ICPR) (pp. 1018–1023). IEEE.
230.Zhang, T., & Guo, G. (2020). Graph attention lstm: A spat-temperal approach for traffic flow forecasting. IEEE Intelligent Transportation Systems Magazine, .
231.Zhang, T., Jin, J., Yang, H., Guo, H., & Ma, X. (2019c). Link speed prediction for signalized urban traffic network using a hybrid deep learning approach. In 2019 IEEE Intelligent Transportation Systems Conference (ITSC) (pp. 2195–2200). IEEE.
232.Zhang, W., Liu, H., Liu, Y., Zhou, J., & Xiong, H. (2020f). Semi-supervised hierarchical recurrent graph neural network for city-wide parking availability prediction. In Proceedings of the AAAI Conference on Artificial Intelligence. volume 34.
233.Zhang, W., Liu, H., Liu, Y., Zhou, J., Xu, T., & Xiong, H. (2020g). Semi-supervised city-wide parking availability prediction via hierarchical recurrent graph neural network. IEEE Transactions on Knowledge and Data Engineering, .
234.Zhang, X., Huang, C., Xu, Y., & Xia, L. (2020h). Spatial-temporal convolutional graph attention networks for citywide traffic flow forecasting. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management (pp. 1853–1862).
235.Zhang, X., Zhang, Z., & Jin, X. (2020i). Spatial-temporal graph attention model on traffic forecasting. In 2020 13th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI) (pp. 999–1003). IEEE.
236.Zhang, Y., Cheng, T., & Ren, Y. (2019d). A graph deep learning method for short-term traffic forecasting on large road networks. Computer-Aided Civil and Infrastructure Engineering, 34 , 877–896.
237.Zhang, Y., Cheng, T., Ren, Y., & Xie, K. (2020j). A novel residual graph convolution deep learning model for short-term network-based traffic forecasting. International Journal of Geographical Information Science, 34 , 969–995.
238.Zhang, Y., Dong, X., Shang, L., Zhang, D., & Wang, D. (2020k). A multi-modal graph neural network approach to traffic risk forecasting in smart urban sensing. In 2020 17th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON) (pp. 1–9). IEEE.
239.Zhang, Y., Lu, M., & Li, H. (2020l). Urban traffic flow forecast based on fastgcrnn. Journal of Advanced Transportation, 2020 .
240.Zhang, Y., Wang, S., Chen, B., & Cao, J. (2019e). Gcgan: Generative adversarial nets with graph cnn for network-scale traffic prediction. In 2019 International Joint Conference on Neural Networks (IJCNN) (pp. 1–8). IEEE.
241.Zhang, Z., Cui, P., & Zhu, W. (2020m). Deep learning on graphs: A survey. IEEE Transactions on Knowledge and Data Engineering, .
242.Zhang, Z., Li, M., Lin, X., Wang, Y., & He, F. (2019f). Multistep speed prediction on traffic networks: A deep learning approach considering spatio-temporal dependencies. Transportation research part C: emerging technologies, 105 , 297–322.
243.Zhao, B., Gao, X., Liu, J., Zhao, J., & Xu, C. (2020a). Spatiotemporal data fusion in graph convolutional networks for traffic prediction. IEEE Access, .
244.Zhao, H., Yang, H., Wang, Y., Wang, D., & Su, R. (2020b). Attention based graph bi-lstm networks for traffic forecasting. In 2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC) (pp. 1–6). IEEE.
245.Zhao, L., Song, Y., Zhang, C., Liu, Y., Wang, P., Lin, T., Deng, M., & Li, H. (2019). T-gcn: A temporal graph convolutional network for traffic prediction. IEEE Transactions on Intelligent Transportation Systems, .
246.Zhao, T., Liu, Y., Neves, L., Woodford, O., Jiang, M., & Shah, N. (2021). Data augmentation for graph neural networks. In Proceedings of the 30th International Joint Conference on Artificial Intelligence. AAAI Press.
247.Zheng, B., Hu, Q., Ming, L., Hu, J., Chen, L., Zheng, K., & Jensen, C. S. (2020a). Spatial-temporal demand forecasting and competitive supply via graph convolutional networks. arXiv preprint arXiv:2009.12157 , .
248.Zheng, C., Fan, X., Wang, C., & Qi, J. (2020b). Gman: A graph multi-attention network for traffic prediction. In Proceedings of the AAAI Conference on Artificial Intelligence. volume 34.
249.Zhou, F., Yang, Q., Zhang, K., Trajcevski, G., Zhong, T., & Khokhar, A. (2020a). Reinforced spatio-temporal attentive graph neural networks for traffic forecasting. IEEE Internet of Things Journal, .
250.Zhou, F., Yang, Q., Zhong, T., Chen, D., & Zhang, N. (2020b). Variational graph neural networks for road traffic prediction in intelligent transportation systems. IEEE Transactions on Industrial Informatics, .
251.Zhou, J., Cui, G., Hu, S., Zhang, Z., Yang, C., Liu, Z., Wang, L., Li, C., & Sun, M. (2020c). Graph neural networks: A review of methods and applications. AI Open, 1 , 57–81.
252.Zhou, Q., Gu, J.-J., Ling, C., Li, W.-B., Zhuang, Y., & Wang, J. (2020d). Exploiting multiple correlations among urban regions for crowd flow prediction. Journal of Computer Science and Technology, 35 , 338–352.
253.Zhou, X., Shen, Y., & Huang, L. (2019). Revisiting flow information for traffic prediction. arXiv preprint arXiv:1906.00560 , .
254.Zhou, Z., Wang, Y., Xie, X., Chen, L., & Liu, H. (2020e). Riskoracle: A minute-level citywide traffic accident forecasting framework. In Proceedings of the AAAI Conference on Artificial Intelligence. volume 34.
255.Zhou, Z., Wang, Y., Xie, X., Chen, L., & Zhu, C. (2020f). Foresee urban sparse traffic accidents: A spatiotemporal multi-granularity perspective. IEEE Transactions on Knowledge and Data Engineering, .
256.Zhu, H., Luo, Y., Liu, Q., Fan, H., Song, T., Yu, C. W., & Du, B. (2019). Multistep flow prediction on car-sharing systems: A multi-graph convolutional neural network with attention mechanism. International Journal of Software Engineering and Knowledge Engineering, 29 , 1727–1740.
257.Zhu, H., Xie, Y., He, W., Sun, C., Zhu, K., Zhou, G., & Ma, N. (2020). A novel traffic flow forecasting method based on rnn-gcn and brb. Journal of Advanced Transportation, 2020 .
258.Zhu, J., Han, X., Deng, H., Tao, C., Zhao, L., Wang, P., Lin, T., & Li, H. (2022). Kst-gcn: A knowledge-driven spatial-temporal graph convolutional network for traffic forecasting. IEEE Transactions on Intelligent Transportation Systems, .
259.Zhu, J., Wang, Q., Tao, C., Deng, H., Zhao, L., & Li, H. (2021). Ast-gcn: Attribute-augmented spatiotemporal graph convolutional network for traffic forecasting. IEEE Access, 9 , 35973–35983.

Citation

MLA

Jiang, W., and J. Luo. “Graph Neural Network for Traffic Forecasting: A Survey”. Expert Systems with Applications, vol. 207, 2022, p. 117921, https://doi.org/10.1016/j.eswa.2022.117921.

APA

Jiang, W., & Luo, J. (2022). Graph neural network for traffic forecasting: A survey. Expert Systems with Applications, 207, 117921. https://doi.org/10.1016/j.eswa.2022.117921

Chicago

Jiang, W., and J. Luo. 2022. “Graph Neural Network for Traffic Forecasting: A Survey”. Expert Systems with Applications 207: 117921. https://doi.org/10.1016/j.eswa.2022.117921.

Harvard

Jiang, W. and Luo, J. (2022) “Graph neural network for traffic forecasting: A survey”, Expert Systems with Applications, 207, p. 117921. Available at: https://doi.org/10.1016/j.eswa.2022.117921.

Vancouver

1. Jiang W, Luo J (2022) Graph neural network for traffic forecasting: A survey. Expert Systems with Applications 207:117921

BibTeX

@article{Jiang_2022, title={Graph neural network for traffic forecasting: A survey}, volume={207}, ISSN={0957-4174}, url={http://dx.doi.org/10.1016/j.eswa.2022.117921}, DOI={10.1016/j.eswa.2022.117921}, journal={Expert Systems with Applications}, publisher={Elsevier BV}, author={Jiang, Weiwei and Luo, Jiayun}, year={2022}, month=Nov, pages={117921} }

Metadata:Crossref

Source Code

This paper has an official code repository available. Click below to access the source code.

View Repository

Access the Paper

This paper is available from its original source. Click below to access the PDF.

Open PDF

License: https://creativecommons.org/licenses/by/4.0/

Graph Neural Network for Traffic Forecasting: A Survey

Abstract

Topics

Keywords

Table of Contents

Knowls

Knowl 1 — Formal Definition of Traffic Graph and Graph-Based Traffic Forecasting

Knowl 2 — Classification of Traffic Forecasting Problems

Knowl 3 — Taxonomy of Adjacency Matrix Construction Methods in Traffic Graphs

Knowl 4 — Spectral Graph Convolution and Renormalized Layer Formulation

Knowl 5 — Diffusion Graph Convolution for Directed Traffic Networks

Knowl 6 — Spatial Message Passing and Graph Attention Operations for Traffic Modeling

Knowl 7 — Taxonomy of Temporal Modeling in Spatiotemporal Graph Neural Networks