Papers › Domain Adversarial Spatial-Temporal Network: A Transferable Framework for Short-term...

Domain Adversarial Spatial-Temporal Network: A Transferable Framework for Short-term Traffic Forecasting across Cities

8 Feb 2022arXiv:2202.03630archive 2025-07-28

Yihong Tang, Ao Qu, Andy H. F. Chow, William H. K. Lam, S. C. Wong, Wei Ma

Accurate real-time traffic forecast is critical for intelligent transportation systems (ITS) and it serves as the cornerstone of various smart mobility applications. Though this research area is dominated by deep learning, recent studies indicate that the accuracy improvement by developing new model structures is becoming marginal. Instead, we envision that the improvement can be achieved by transferring the "forecasting-related knowledge" across cities with different data distributions and network topologies. To this end, this paper aims to propose a novel transferable traffic forecasting framework: Domain Adversarial Spatial-Temporal Network (DASTNet). DASTNet is pre-trained on multiple source networks and fine-tuned with the target network's traffic data. Specifically, we leverage the graph representation learning and adversarial domain adaptation techniques to learn the domain-invariant node embeddings, which are further incorporated to model the temporal traffic data. To the best of our knowledge, we are the first to employ adversarial multi-domain adaptation for network-wide traffic forecasting problems. DASTNet consistently outperforms all state-of-the-art baseline methods on three benchmark datasets. The trained DASTNet is applied to Hong Kong's new traffic detectors, and accurate traffic predictions can be delivered immediately (within one day) when the detector is available. Overall, this study suggests an alternative to enhance the traffic forecasting methods and provides practical implications for cities lacking historical traffic data.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

yihongt/dastnet officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Domain AdaptationGraph MiningGraph Representation LearningRepresentation LearningSpatio-Temporal ForecastingTime Series ForecastingTime Series RegressionTraffic PredictionTransfer Learning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Traffic Prediction PeMSD4 (10 days' training data, 15min) DASTNet MAE 19.25 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD4 (10 days' training data, 15min) DASTNet MAPE 13.30 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD4 (10 days' training data, 15min) DASTNet RMSE 28.91 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD4 (10 days' training data, 30min) DASTNet MAE 20.67 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD4 (10 days' training data, 30min) DASTNet MAPE 14.56 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD4 (10 days' training data, 30min) DASTNet RMSE 30.78 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD4 (10 days' training data, 60min) DASTNet MAE 22.82 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD4 (10 days' training data, 60min) DASTNet MAPE 16.1 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD4 (10 days' training data, 60min) DASTNet RMSE 33.77 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD7 (10 days' training data, 15min) DASTNet MAE 20.91 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD7 (10 days' training data, 15min) DASTNet MAPE 8.95 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD7 (10 days' training data, 15min) DASTNet RMSE 31.85 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD7 (10 days' training data, 30min) DASTNet MAE 22.96 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD7 (10 days' training data, 30min) DASTNet MAPE 9.87 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD7 (10 days' training data, 30min) DASTNet RMSE 34.8 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD7 (10 days' training data, 60min) DASTNet MAE 26.88 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD7 (10 days' training data, 60min) DASTNet MAPE 11.75 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD7 (10 days' training data, 60min) DASTNet RMSE 40.12 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD8 (10 days' training data, 15min) DASTNet MAE 15.26 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD8 (10 days' training data, 15min) DASTNet MAPE 9.64 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD8 (10 days' training data, 15min) DASTNet RMSE 22.7 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD8 (10 days' training data, 30min) DASTNet MAE 16.41 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD8 (10 days' training data, 30min) DASTNet MAPE 10.46 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD8 (10 days' training data, 30min) DASTNet RMSE 24.57 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD8 (10 days' training data, 60min) DASTNet MAE 18.84 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD8 (10 days' training data, 60min) DASTNet MAPE 11.72 #1 of 1 Archive leaderboard report
Traffic Prediction PeMSD8 (10 days' training data, 60min) DASTNet RMSE 28.06 #1 of 1 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

GINGRUnode2vec

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections