Research map / Probabilistic and scientific modeling / Causal discovery
Local Causal Discovery With Background Knowledge
Q. Zheng, Yue Liu, Yangbo He
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 LABITER local causal identification instead of whole-MPDAG learning (FairRelax -> LFairRelax) | relaxed fair predictor · counterfactual fairness, 10 nodes | unfairness: 0.069 -> 0.012 | Improved |
| use LABITER instead of Algorithm 2 for Fair predictor | Fair predictor · counterfactual fairness, 10 nodes | unfairness: 0.058 -> 0.012 | Improved |
| use LABITER instead of Algorithm 2 for Fair predictor | Fair predictor · counterfactual fairness, 40 nodes | unfairness: 0.026 -> 0.025 | Improved |
| use LABITER for Fair predictor | Fair predictor · counterfactual fairness, 40 nodes | RMSE: 0.843 -> 0.947 | Hurt |
| use fair predictor (LFair) instead of Full model | Full predictor · counterfactual fairness, 10 nodes | unfairness: 0.251 -> 0.012 | Improved |
| use fair predictor (LFair) instead of Full model | Full predictor · counterfactual fairness, 10 nodes | RMSE: 0.431 -> 0.59 | Hurt |
| restrict to definite non-descendants only instead of also including possible descendants (LFairRelax vs LFair) (ablation) | LABITER predictors · counterfactual fairness, 20 nodes | unfairness: 0.035 -> 0.034 | No clear effect |
| increase proportion of background knowledge from 0.1 to 0.9 (ablation) | MB-by-MB in MPDAG · local structure learning | number of CI tests | Improved |
| add background knowledge to MB-by-MB in MPDAG vs MB-by-MB and PC | MB-by-MB in MPDAG · local structure learning, many known edges | SHD | Improved |
| use LABITER instead of local ITC and Algorithm 2 to identify causal relations | LABITER · random graphs n=50 and 100 | Kappa coefficient | Improved |
| use LABITER on Bank dataset vs Algorithm 2 with learned MPDAG | LABITER · Bank marketing, marital status | definite descendants | Mixed |
Closest papers by meaning
- Local Identifying Causal Relations in the Presence of Latent Variables (ICML 2025 · similarity 0.93)
- Integrating Background Knowledge for Scalable Causal Discovery (NeurIPS 2026 · similarity 0.92)
- Local Causal Structure Learning in the Presence of Latent Variables (ICML 2024 · similarity 0.90)
- Local Causal Discovery Without Causal Sufficiency (AAAI 2025 · similarity 0.90)
- THE ROBUSTNESS OF DIFFERENTIABLE CAUSAL DISCOVERY IN MISSPECIFIED SCENARIOS (ICLR 2025 · similarity 0.89)
- CausalTab: Pretraining Across Causal Environments for Tabular Causal Discovery (NeurIPS 2026 · 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.
