Research map / Probabilistic and scientific modeling / Causal discovery
Identifiability of Additive Noise Models Using Conditional Variances
Gunwoong Park
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 forward/backward stepwise ordering (USF1/USB1) instead of greedy search GDS | DAG learning on mathematics marks · Mathematics marks (88 students, 5 subjects) | log-likelihood: -1711.135 -> -1695.589 | Improved |
| use forward/backward stepwise ordering (USF1) instead of LISTEN | DAG learning on mathematics marks · Mathematics marks (88 students, 5 subjects) | BIC: -1793.848 -> -1731.407 | Improved |
| add oracle true skeleton to estimated ordering (USF1Oracle vs USF1) (ablation) | US algorithm, Gaussian linear SEM simulations · Simulated Gaussian linear SEMs, homogeneous variance | Hamming distance | Improved |
| forward selection (USF1Oracle) versus backward selection (USB1Oracle) ordering (ablation) | US algorithm, p=200 heterogeneous variances · Simulated Gaussian linear SEMs, heterogeneous variance | ordering recovery | Improved |
| use quadratic regression forward selection (USF2) instead of backward (USB2) (ablation) | US algorithm, 20-node polynomial SEMs · Simulated Gaussian polynomial SEMs | Hamming distance | Improved |
| increase graph density from d=2 to d=4 (ablation) | USF1/USB1 · Simulated Gaussian linear SEMs, homogeneous variance | precision/recall/Hamming | Hurt |
| violate identifiability condition with error variances 3, 0.25, 1.25 (ablation) | USF1/USB1/LISTEN · Simulated 20-node linear SEMs, non-Gaussian errors | graph recovery: error variances sequentially 3, 0.25, and 1.25 | Hurt |
| increase sample size n from 250 to 2500 (ablation) | LINGAM on Gaussian linear SEMs · Simulated Gaussian linear SEMs, homogeneous variance | accuracy | Hurt |
Closest papers by meaning
- Identifiability of Linear AMP Chain Graph Models (AAAI 2022 · similarity 0.86)
- Strong and Weak Identifiability of Optimization-based Causal Discovery in Non-linear Additive Noise Models (ICML 2025 · similarity 0.85)
- Achievable distributional robustness when the robust risk is only partially identified (NeurIPS 2024 · similarity 0.84)
- Effective Causal Discovery under Identifiable Heteroscedastic Noise Model (AAAI 2024 · similarity 0.84)
- Variance estimation in compound decision theory under boundedness (NeurIPS 2024 · similarity 0.84)
- Identifiability Analysis of Linear ODE Systems with Hidden Confounders (NeurIPS 2024 · similarity 0.83)
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.
