Research map / Graph learning / Knowledge and temporal graphs
Generic and Dynamic Graph Representation Learning for Crowd Flow Modeling
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 GDCF instead of XGBoost for inflow prediction | BJSubway · BJSubway inflow | MAE: 103.9266 -> 78.9725 | Helped |
| Use GDCF instead of XGBoost for outflow prediction | BJSubway · BJSubway outflow | MAE: 121.0290 -> 71.7893 | Helped |
| Use GDCF instead of DNEAT for OD flow prediction | BJSubway · BJSubway OD flow | MAE: 1.4772 -> 1.3926 | Helped |
| Use GDCF instead of XGBoost for inflow prediction | NYTaxi · NYTaxi inflow | MAE: 10.7097 -> 10.5956 | Helped |
| Use GDCF instead of DyRep for outflow prediction | NYTaxi · NYTaxi outflow | MAE: 9.7327 -> 9.4054 | Helped |
| Use GDCF instead of DyRep for OD flow prediction | NYTaxi · NYTaxi OD flow | MAE: 0.6049 -> 0.5807 | Helped |
| Add encoder-decoder reconstruction pretraining rather than training end-to-end (ablation) | GDCF on BJSubway crowd flow prediction · BJSubway crowd flow prediction | MAE | Helped |
| Add adaptive global relation learning to propagate information among traffic nodes (ablation) | GDCF on BJSubway crowd flow prediction · BJSubway crowd flow prediction | MAE | Helped |
| Use learned GDCF node representations instead of OD matrix representations | NYTaxi travel time estimation · NYTaxi travel time estimation | MAE: 156.0614 -> 141.7542 | Helped |
| Use phase-1 learned node representations and train the downstream model with 7 days rather than 42 days of data | GDCF on BJSubway · BJSubway crowd flow prediction | MAE | No clear effect |
Closest papers by meaning
- Explainable Origin-Destination Crowd Flow Interpolation via Variational Multi-Modal Recurrent Graph Auto-Encoder (AAAI 2024 · similarity 0.87)
- Deep Spatio-Temporal Residual Networks for Citywide Crowd Flows Prediction (AAAI 2017 · similarity 0.86)
- DSTAGNN: Dynamic Spatial-Temporal Aware Graph Neural Network for Traffic Flow Forecasting (ICML 2022 · similarity 0.85)
- Towards Dynamic Spatial-Temporal Graph Learning: A Decoupled Perspective (AAAI 2024 · similarity 0.85)
- Efficient Traffic Prediction Through Spatio-Temporal Distillation (AAAI 2025 · similarity 0.82)
- Spatiotemporal-aware Trend-Seasonality Decomposition Network for Traffic Flow Forecasting (AAAI 2025 · similarity 0.82)
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.