Research map / Graph learning / Knowledge and temporal graphs
Environment-Aware Dynamic Graph Learning for Out-of-Distribution Generalization
Haonan Yuan, Qingyun Sun, Xingcheng Fu, Ziwei Zhang, Cheng Ji, Hao Peng, Jianxin Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 EAGLE instead of DIDA | future link prediction · COLLAB w/ OOD | AUC: 81.87±0.40 -> 84.41±0.87 | Improved |
| use EAGLE instead of DIDA | future link prediction · Yelp w/ OOD | AUC: 75.92±0.90 -> 77.26±0.74 | Improved |
| use EAGLE instead of DIDA | future link prediction · ACT w/ OOD | AUC: 78.64±0.97 -> 82.70±0.72 | Improved |
| use EAGLE instead of DIDA | future link prediction · COLLAB w/o OOD | AUC: 91.97±0.05 -> 92.45±0.21 | Improved |
| use EAGLE instead of GroupDRO | future link prediction · Yelp w/o OOD | AUC: 79.38±0.42 -> 78.97±0.31 | Hurt |
| use EAGLE instead of DIDA | future link prediction · ACT w/o OOD | AUC: 89.84±0.82 -> 92.37±0.53 | Improved |
| use EAGLE instead of DIDA | future link prediction with node feature shifts · COLLAB node feature shift, p̄=0.4 test | AUC: 85.20±0.84 -> 88.32±0.61 | Improved |
| use EAGLE instead of DIDA | future link prediction with node feature shifts · COLLAB node feature shift, p̄=0.6 test | AUC: 82.89±0.23 -> 87.29±0.71 | Improved |
| use EAGLE instead of DIDA | future link prediction with node feature shifts · COLLAB node feature shift, p̄=0.8 test | AUC: 72.59±3.31 -> 82.30±0.75 | Improved |
| add environment instantiation and generated environment samples (ablation) | EAGLE · all ablation-study datasets | performance | Improved |
| add invariant pattern recognition instead of randomly generating node thresholds (ablation) | EAGLE · all ablation-study datasets | performance | Improved |
| add the spatio-temporal causal intervention mechanism and risk loss (ablation) | EAGLE · all ablation-study datasets | performance | Improved |
Closest papers by meaning
- Improving Generalization of Dynamic Graph Learning via Environment Prompt (NeurIPS 2024 · similarity 0.92)
- Improving Out-of-Distribution Generalization in Graphs via Hierarchical Semantic Environments (CVPR 2024 · similarity 0.90)
- Expand Horizon: Graph Out-of-Distribution Generalization via Multi-Level Environment Inference (AAAI 2025 · similarity 0.90)
- Graph Invariant Learning with Subgraph Co-mixup for Out-of-Distribution Generalization (AAAI 2024 · similarity 0.89)
- Topology-Aware Dynamic Reweighting for Distribution Shifts on Graph (ICML 2025 · similarity 0.88)
- Towards Dynamic Spatial-Temporal Graph Learning: A Decoupled Perspective (AAAI 2024 · 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.
