Research map / Graph learning / Knowledge and temporal graphs
Learning Geo-Contextual Embeddings for Commuting Flow Prediction
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 |
|---|---|---|---|
| add multitask in/out-flow subtasks (ablation) | GMEL · NYC commuting flow test set | RMSE: 5.356 -> 4.887 | Helped |
| add multitask in/out-flow subtasks (ablation) | GMEL · NYC commuting flow test set | CPC: 0.716 -> 0.741 | Helped |
| use separate origin and destination GATs instead of one shared GAT (ablation) | GMEL · NYC commuting flow test set | RMSE: 4.982 -> 4.887 | Helped |
| use separate origin and destination GATs instead of one shared GAT (ablation) | GMEL · NYC commuting flow test set | CPC: 0.737 -> 0.741 | Helped |
| use GMEL geo-contextual embeddings as GBRT input instead of raw features | GBRT · NYC commuting flow test set | RMSE: 5.454 -> 4.887 | Helped |
| use Node2vec embeddings instead of raw features for GBRT | GBRT · NYC commuting flow test set | RMSE: 5.454 -> 5.455 | No clear effect |
| use GBRT instead of gravity model | flow predictor · NYC commuting flow test set | RMSE: 6.944 -> 5.454 | Helped |
| use Random Forest instead of gravity model | flow predictor · NYC commuting flow test set | MAE: 2.179 -> 2.436 | Hurt |
| use 1 GAT layer instead of 2 or more (ablation) | GMEL · NYC commuting flow test set | prediction error | Hurt |
| increase embedding size from 32 to 128 (ablation) | GMEL · NYC commuting flow test set | prediction error | Helped |
Closest papers by meaning
- GeoPro-Net: Learning Interpretable Spatiotemporal Prediction Models Through Statistically-Guided Geo-Prototyping (AAAI 2025 · similarity 0.85)
- Efficient Traffic Prediction Through Spatio-Temporal Distillation (AAAI 2025 · similarity 0.85)
- Satellites Reveal Mobility: A Commuting Origin-destination Flow Generator for Global Cities (NeurIPS 2025 · similarity 0.85)
- Bridging Traffic State and Trajectory for Dynamic Road Network and Trajectory Representation Learning (AAAI 2025 · similarity 0.85)
- UniTR: A Unified Framework for Joint Representation Learning of Trajectories and Road Networks (AAAI 2025 · similarity 0.85)
- FlashST: A Simple and Universal Prompt-Tuning Framework for Traffic Prediction (ICML 2024 · similarity 0.84)
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.