Research map / Probabilistic and scientific modeling / Causal discovery
Spatio-Temporal Hierarchical Causal Models
Xintong Li, Haoran Zhang, Xiao Zhou
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 |
|---|---|---|---|
| model unit-level hierarchy (T-HCM) instead of aggregating | Aggregated LMM · Synthetic (rho=2.0, gamma=4.0) | ATE absolute error: 7.94 vs an order of magnitude smaller | Improved |
| add spatial modeling (ST-HCM) to T-HCM | T-HCM · Synthetic (gamma=4.0, rho 0.0 to 2.0) | ATE absolute error: 0.11 to 2.35 for T-HCM vs about 0.11 for ST-HCM | Improved |
| add spatial modeling (ST-HCM) to T-HCM when no spillover | T-HCM · Synthetic (rho=0.0, gamma=0.0) | ATE absolute error: 0.054 -> 0.057 | No clear effect |
| add spatial modeling (ST-HCM) to T-HCM | T-HCM · Synthetic (rho=1.0, gamma=0.0) | ATE absolute error: 0.132 -> 0.056 | Improved |
| increase number of subunits per unit m | ST-HCM estimator · Synthetic canonical process | absolute error | Improved |
| relax time-invariance of confounder (temporal drift) (ablation) | ST-HCM (LMM) · Synthetic robustness (N=16, m=50, T=8) | ATE error | Hurt |
| violate spatial ordering with feedback loops (ablation) | ST-HCM (LMM) · Synthetic robustness (N=16, m=50, T=8) | ATE error | No clear effect |
| add hierarchy then spatial structure to aggregated model (ablation) | Aggregated model · Chicago traffic | estimated ATE | Mixed |
Closest papers by meaning
- Discovering Latent Causal Graphs from Spatiotemporal Data (ICML 2025 · similarity 0.91)
- Causal Deciphering and Inpainting in Spatio-Temporal Dynamics via Diffusion Model (NeurIPS 2024 · similarity 0.90)
- Causal Spatio-Temporal Prediction: An Effective and Efficient Multi-Modal Approach (NeurIPS 2025 · similarity 0.90)
- SPACETIME: Causal Discovery from Non-Stationary Time Series (AAAI 2025 · similarity 0.89)
- GST-UNet: A Neural Framework for Spatiotemporal Causal Inference with Time-Varying Confounding (NeurIPS 2025 · similarity 0.88)
- CausalRivers - Scaling up benchmarking of causal discovery for real-world time-series (ICLR 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.
