Research map / Video / Video segmentation and tracking
ST-HHOL: Spatio-Temporal Hierarchical Hypergraph Online Learning for Crime Prediction
Keqing Du, Yufan Kang, Xinyu Yang, Wei Shao
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 |
|---|---|---|---|
| add heterogeneous hypergraph G^e with multi-source inputs (ablation) | ST-HHOL · four crime datasets | MAE/RMSE/MAPE | Improved |
| add homogeneous co-occurrence hypergraph G^o (ablation) | ST-HHOL · four crime datasets | MAE/RMSE/MAPE | Improved |
| add partially frozen LLM (PF-LLM) instead of standard Transformer (ablation) | ST-HHOL · four crime datasets | error | Improved |
| add online learning strategy (ablation) | ST-HHOL · four crime datasets | error | Improved |
| freeze FFN layers only (PF-FFN) instead of full tuning (ablation) | PF-LLM · four crime datasets | MAE/RMSE | Improved |
| fine-tune every 0.5 month instead of every 0.25 month (ablation) | ST-HHOL · CHI | MAE/RMSE/MAPE | Improved |
| fine-tune every 0.75 month instead of 0.5 month under 3-month retraining (ablation) | ST-HHOL · CHI | MAE | Mixed |
| retrain every two months instead of other intervals (ablation) | ST-HHOL · CHI | MAE/RMSE/MAPE | Improved |
| use 64 hyperedges instead of 128 (ablation) | ST-HHOL · CHI | MAE/RMSE/MAPE | Improved |
| use hidden dimension 16 instead of larger dimensions (ablation) | ST-HHOL · CHI | MAE | Improved |
| use 2 partially frozen layers instead of more than two (ablation) | PF-LLM · CHI | error | Improved |
| use batch size 32 (ablation) | ST-HHOL · CHI | error | Improved |
Closest papers by meaning
- Hawkes-Enhanced Spatial-Temporal Hypergraph Contrastive Learning Based on Criminal Correlations (AAAI 2024 · similarity 0.89)
- Spatial-Temporal Augmentation for Crime Prediction (Student Abstract) (AAAI 2024 · similarity 0.89)
- GeoPro-Net: Learning Interpretable Spatiotemporal Prediction Models Through Statistically-Guided Geo-Prototyping (AAAI 2025 · similarity 0.83)
- Towards Dynamic Spatial-Temporal Graph Learning: A Decoupled Perspective (AAAI 2024 · similarity 0.82)
- Alleviating Performance Disparity in Adversarial Spatiotemporal Graph Learning Under Zero-Inflated Distribution (AAAI 2025 · similarity 0.81)
- STRAP: Spatio-Temporal Pattern Retrieval for Out-of-Distribution Generalization (NeurIPS 2025 · similarity 0.81)
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.
