Research map / Probabilistic and scientific modeling / Variational and generative sampling
A Gibbs Sampler for Learning DAGs
Robert J. B. Goudie, Sach Mukherjee
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 Gibbs sampler instead of MC3 | structure MCMC · ALARM (n=1000) | max log marginal likelihood: −11,176.9 ± 273.7 -> −10,608.20 ± 0.8 | Improved |
| use Gibbs sampler instead of REV | structure MCMC · ALARM (n=1000) | max log marginal likelihood: −10,627.2 ± 40.4 -> −10,608.20 ± 0.8 | Improved |
| use Gibbs sampler instead of REV | structure MCMC · ALARM, FPR 0 (n=5000) | true edges found: 13.2 ± 21.3 -> 43.0 ± 0.0 | Improved |
| use Gibbs sampler instead of MC3 | structure MCMC · ALARM, FPR 0 (n=5000) | true edges found: 0 -> 43.0 ± 0.0 | Improved |
| use Gibbs sampler instead of MC3 | structure MCMC · ALARM, FPR 0 (n=100) | true edges found: 21.7 ± 2.1 -> 21.9 ± 1.9 | No clear effect |
| increase sample size from n=100 to 5000 | MC3 · ALARM, FPR 0 | true edges found: 21.7 ± 2.1 -> 0 | Hurt |
| increase sample size from n=100 to 5000 | REV · ALARM, FPR 0 | true edges found: 20.1 ± 1.4 -> 13.2 ± 21.3 | Hurt |
| increase sample size from n=100 to 5000 | Gibbs sampler · ALARM, FPR 0 | true edges found: 21.9 ± 1.9 -> 43.0 ± 0.0 | Improved |
| use Gibbs sampler instead of MC3 | structure MCMC · Single-cell data | edges differing by 0.1 or more between runs: 151.8 -> 6.4 | Improved |
| use Gibbs sampler instead of REV | structure MCMC · Single-cell data | major between-run discrepancies: 26 -> 0 | Improved |
| use Gibbs sampler instead of PC | structure estimator · Single-cell bootstrap | mean SHD between bootstrap estimates: 214.6 -> 128.8 | Improved |
| use Gibbs sampler instead of MC3 | structure MCMC · Survey data BRFSS | major between-run discrepancies: 10.9 -> 0 | Improved |
Closest papers by meaning
- Structure Learning with Adaptive Random Neighborhood Informed MCMC (NeurIPS 2023 · similarity 0.92)
- Learning Gaussian DAGs from Network Data (JMLR 2024 · similarity 0.91)
- ProDAG: Projected Variational Inference for Directed Acyclic Graphs (NeurIPS 2025 · similarity 0.89)
- Markov Equivalence and Consistency in Differentiable Structure Learning (NeurIPS 2024 · similarity 0.89)
- Boosting Causal Structure Learning: An Asymmetric Exponential Modulation Gaussian-Based Adaptive Sample Reweighting Framework (AAAI 2025 · similarity 0.88)
- Distribution Learning Meets Graph Structure Sampling (NeurIPS 2025 · similarity 0.88)
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.
