Research map / Graph learning / Knowledge and temporal graphs
Spatial-Temporal Fusion Graph Neural Networks for Traffic Flow Forecasting
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use STFGNN instead of the strongest listed baseline, STSGCN | Traffic forecasting · PEMS03 | MAE: 17.48 ± 0.15 -> 16.77 ± 0.09 | Helped |
| Use STFGNN instead of STSGCN | Traffic forecasting · PEMS04 | MAE: 21.19 ± 0.10 -> 19.83± 0.06 | Helped |
| Use STFGNN instead of STSGCN | Traffic forecasting · PEMS07 | MAE: 24.26 ± 0.14 -> 22.07± 0.11 | Helped |
| Use STFGNN instead of STSGCN | Traffic forecasting · PEMS08 | MAE: 17.13 ± 0.09 -> 16.64± 0.09 | Helped |
| Increase fusion graph size from 3 to 4 (ablation) | STFGNN without gated convolution, temporal graph sparsity about 1% · PEMS04 | MAE: 20.09 -> 19.92 | Helped |
| Increase fusion graph size from 3 to 4 (ablation) | STFGNN without gated convolution, temporal graph sparsity about 1% · PEMS08 | MAE: 16.84 -> 16.70 | Helped |
| Add gated convolution (ablation) | STFGNN with fusion graph size 4 and temporal graph sparsity about 1% · PEMS04 | MAE: 19.92 -> 19.83 | Helped |
| Add gated convolution (ablation) | STFGNN with fusion graph size 4 and temporal graph sparsity about 1% · PEMS08 | MAE: 16.70 -> 16.64 | No clear effect |
| Replace the spatial graph with the temporal graph (ablation) | STFGNN, fusion graph size 4 and sparsity about 1% · PEMS04 | MAE: 19.83 -> 20.02 | Hurt |
| Replace the spatial graph with the temporal graph (ablation) | STFGNN, fusion graph size 4 and sparsity about 1% · PEMS08 | MAE: 16.64 -> 18.23 | Hurt |
| Increase temporal graph sparsity from about 1% to about 5% in temporal-only configuration (ablation) | STFGNN with fusion graph size 4 and gated convolution · PEMS04 | MAE: 20.02 -> 19.91 | No clear effect |
| Increase temporal graph sparsity from about 1% to about 5% in temporal-only configuration (ablation) | STFGNN with fusion graph size 4 and gated convolution · PEMS08 | MAE: 18.23 -> 16.02 | Helped |
Closest papers by meaning
- Spatio-Temporal Pivotal Graph Neural Networks for Traffic Flow Forecasting (AAAI 2024 · similarity 0.92)
- Efficient Traffic Prediction Through Spatio-Temporal Distillation (AAAI 2025 · similarity 0.90)
- Spatiotemporal-aware Trend-Seasonality Decomposition Network for Traffic Flow Forecasting (AAAI 2025 · similarity 0.89)
- Towards Dynamic Spatial-Temporal Graph Learning: A Decoupled Perspective (AAAI 2024 · similarity 0.89)
- Disentanglement-Guided Spatial-Temporal Graph Neural Network for Metro Flow Forecasting (Student Abstract) (AAAI 2024 · similarity 0.88)
- Designing Specialized Two-Dimensional Graph Spectral Filters for Spatial-Temporal Graph Modeling (AAAI 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.