Research map / Probabilistic and scientific modeling / Causal discovery
How to Gain on Power: Novel Conditional Independence Tests Based on Short Expansion of Conditional Mutual Information
Mariusz Kubkowski, Jan Mielniczuk, Paweł Teisseyre
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 SECMI test instead of asymptotic CMI test in Grow-Shrink | GS · MB discovery sachs (n=1000) | F-measure: 0.651 -> 0.932 | Improved |
| use SECMI test instead of asymptotic CMI test in Grow-Shrink | GS · MB discovery insurance (n=1000) | F-measure: 0.479 -> 0.819 | Improved |
| use SECMI instead of CMI(sp) in Grow-Shrink | GS · MB discovery, 10 networks (n=1000) | average rank (lower better): 2.0 -> 1.6 | No clear effect |
| use SECMI instead of CMI(sp) in Grow-Shrink | GS · MB discovery, 10 networks (n=500) | average rank (lower better): 1.6 -> 2.0 | Hurt |
| include third-order terms (SECMI3 instead of SECMI) (ablation) | GS · MB discovery, 10 networks (n=1000) | average rank (lower better): 1.6 -> 2.8 | Hurt |
| use SECMI or CMI(sp) instead of asymptotic CMI | GS · MB discovery, 10 networks (n=1000) | average rank (lower better): 3.6 -> 1.6 | Improved |
| use SECMI instead of CMI on model P1-P3 power with growing conditioning set size | CI test · simulation models P1-P3 | power | Improved |
| use SECMI (second-order expansion) on model P4 with third-order interaction | SECMI · simulation model P4 | power | Hurt |
| add third-order term (SECMI3 instead of SECMI) on model P4 (ablation) | SECMI · simulation model P4 | power | Improved |
| use SECMI with permutation estimate on large m and small n (m=10, n=1000) | SECMI · simulation models E1, E3 | type I error | Hurt |
Closest papers by meaning
- On the Hardness of Conditional Independence Testing In Practice (NeurIPS 2025 · similarity 0.87)
- A Conditional Independence Test in the Presence of Discretization (ICLR 2025 · similarity 0.86)
- A Sample Efficient Conditional Independence Test in the Presence of Discretization (ICML 2025 · similarity 0.85)
- On the sample complexity of conditional independence testing with Von Mises estimator with application to causal discovery (ICML 2024 · similarity 0.84)
- Testing Causal Models with Hidden Variables in Polynomial Delay via Conditional Independencies (AAAI 2025 · similarity 0.84)
- PAIR-CI: Calibrated Conditional Independence Testing for Causal Discovery with Incomplete Data (NeurIPS 2026 · similarity 0.84)
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.
