Research map / Probabilistic and scientific modeling / Causal discovery
Learning linear non-Gaussian directed acyclic graph with diverging number of nodes
Ruixuan Zhao, Xin He, Junhui Wang
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 TL (topological layer) method instead of Pairwise | DAG learning · Example 1, (n,p)=(200,100) | MCC: 0.5551 -> 0.9252 | Improved |
| use TL instead of MMHC | DAG learning · Example 2, (n,p)=(400,200) | MCC: 0.4749 -> 0.7504 | Improved |
| use TL instead of ICA on TPR | DAG learning · Example 1, (n,p)=(200,100) and (400,200) | TPR: 0.8657 vs Pairwise 0.9564; 0.8543 vs Pairwise 0.9819 | Hurt |
| use TL instead of ICA on TPR | DAG learning · Example 2, (n,p)=(200,100) | TPR: 0.6419 vs ICA 0.7202 | Hurt |
| use TL instead of Pairwise on FDR | DAG learning · Example 1, (n,p)=(200,200) | FDR: 0.7982 -> 0.0099 | Improved |
| use TL instead of PC on relative Frobenius error | DAG learning · Example 2, (n,p)=(400,1000) | rel-Fnorm: 0.9754 -> 0.6752 | Improved |
| increase dimension from p=200 to p=1000 at n=400 | TL · Example 1 | MCC: 0.9237 -> 0.9813 | Improved |
| increase dimension from p=200 to p=1000 at n=400 | TL · Example 2 | MCC: 0.7504 -> 0.7313 | No clear effect |
| use TL instead of MDirect | DAG learning · Example 1, (n,p)=(200,100) | MCC: -0.0011 -> 0.9252 | Improved |
Closest papers by meaning
- Learning Gaussian DAGs from Network Data (JMLR 2024 · similarity 0.90)
- Learning Gaussian DAG Models without Condition Number Bounds (ICML 2025 · similarity 0.87)
- Deep Gaussian Processes on Directed Acyclic Graphs (NeurIPS 2026 · similarity 0.87)
- ProDAG: Projected Variational Inference for Directed Acyclic Graphs (NeurIPS 2025 · similarity 0.87)
- Effective Causal Discovery under Identifiable Heteroscedastic Noise Model (AAAI 2024 · similarity 0.86)
- Coarsening Linear Non-Gaussian Causal Models with Cycles (NeurIPS 2026 · similarity 0.86)
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.
