Research map / Probabilistic and scientific modeling / Causal discovery
Coarse-to-Fine Learning of Dynamic Causal Structures
Dezhi Yang, Qiaoyu Tan, Carlotta Domeniconi, Jun Wang, Lizhen Cui, Guoxian Yu
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 linear interpolation of coarse causal matrices (ablation) | DyCasual · Dynamic synthetic, 10 nodes, W_1 | F1 (%): 72.98 -> 91.42 | Improved |
| add linear interpolation of coarse causal matrices (ablation) | DyCasual · Dynamic synthetic, 20 nodes | F1 (%): W_1: 92.18 -> 92.26; W_25: 91.28 -> 93.64 | Mixed |
| use fully dynamic causal model instead of static NOTEARS-style baselines | DyCasual vs DYNO/NTS-NO · Dynamic synthetic, 20 nodes, W_1 | F1 (%): 3.52 -> 91.43 | Improved |
| use coarse-to-fine encoding instead of per-step ODE graph (DYCAST) | DyCasual · Dynamic synthetic, 80 nodes | F1 (%): DYCAST fails to converge at 80 nodes; DyCasual 95.20 at W_25 | Improved |
| apply method to static linear SEM data vs strongest baseline DYNO | DyCasual · Static linear SEM, 80 nodes | SHD: 17.60 -> 9.00 | Improved |
| apply method to static nonlinear SEM vs NTS-NO | DyCasual · Static nonlinear SEM, 10 nodes | TPR (%): 93.33 -> 92.35 | No clear effect |
| use window causality instead of summary causality (vs CUTS+) | DyCasual · CausalTime Medical | F1 (%): 30.23 -> 56.86 | Improved |
| scale acyclic constraint by 1-norm (h_norm) vs h_log (ablation) | DyCasual · synthetic matrices | stability/runtime | Improved |
| increase noise strength to 0.5 (smaller) (ablation) | DyCasual · Dynamic synthetic | TPR/F1 | Hurt |
| overestimate maximum lag when true lag is 3 (ablation) | DyCasual · Dynamic synthetic, unknown lag | TPR | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- CUTS+ — CUTS: Neural Causal Discovery from Irregular Time-Series Data (ICLR 2023) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- UnCLe: Towards Scalable Dynamic Causal Discovery in Non-linear Temporal Systems (NeurIPS 2025 · similarity 0.90)
- CausalDynamics: A large‐scale benchmark for structural discovery of dynamical causal models (NeurIPS 2025 · similarity 0.90)
- MOCHA: Discovering Multi-Order Dynamic Causal Structure in Temporal Point Processes (NeurIPS 2026 · similarity 0.90)
- CausalTab: Pretraining Across Causal Environments for Tabular Causal Discovery (NeurIPS 2026 · similarity 0.89)
- Discovering Mixtures of Structural Causal Models from Time Series Data (ICML 2024 · similarity 0.89)
- Learning Causal Relations from Subsampled Time Series with Two Time-Slices (ICML 2024 · similarity 0.89)
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.
