Research map / Probabilistic and scientific modeling / Molecules and proteins
Learning to Discover Sparse Graphical Models
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| learn the edge estimator from synthetic sparse-graph samples instead of relying on penalized likelihood | DeepGraph-39 · Gaussian random graphs (n=35,p=39) | Prec@5%: 0.441 $\pm$ 0.011 -> 0.463 $\pm$ 0.009 | Helped |
| learn the edge estimator from synthetic sparse-graph samples instead of relying on penalized likelihood | DeepGraph-39 · Gaussian random graphs (n=100,p=39) | Prec@5%: 0.648 $\pm$ 0.012 -> 0.567 $\pm$ 0.009 | Hurt |
| learn the edge estimator from synthetic sparse-graph samples instead of relying on penalized likelihood | DeepGraph-39 · Gaussian random graphs (n=15,p=39) | Prec@5%: 0.261 $\pm$ 0.009 -> 0.326 $\pm$ 0.009 | Helped |
| learn the edge estimator from synthetic sparse-graph samples instead of relying on penalized likelihood | DeepGraph-39 · Laplace random graphs (n=35,p=39) | Prec@5%: 0.298 $\pm$ 0.009 -> 0.415 $\pm$ 0.010 | Helped |
| learn the edge estimator from synthetic sparse-graph samples instead of relying on penalized likelihood | DeepGraph-39 · Gaussian small-world graphs (n=35,p=39) | Prec@5%: 0.428 $\pm$ 0.007 -> 0.479 $\pm$ 0.007 | Helped |
| average predictions across input permutations (ablation) | DeepGraph-39 · Gaussian random graphs (n=35,p=39) | Prec@5%: 0.463 $\pm$ 0.009 -> \bm{$0.487 \pm 0.010$} | Helped |
| average predictions across input permutations (ablation) | DeepGraph-39 · Gaussian random graphs (n=100,p=39) | Prec@5%: 0.567 $\pm$ 0.009 -> $0.581 \pm 0.008$ | Helped |
| average predictions across input permutations (ablation) | DeepGraph-39 · Gaussian random graphs (n=15,p=39) | Prec@5%: 0.326 $\pm$ 0.009 -> \bm{$0.360 \pm 0.010$} | Helped |
| average predictions across input permutations (ablation) | DeepGraph-39 · Laplace random graphs (n=35,p=39) | Prec@5%: 0.415 $\pm$ 0.010 -> \bm{$0.445 \pm 0.011$} | Helped |
| average predictions across input permutations (ablation) | DeepGraph-39 · Gaussian small-world graphs (n=35,p=39) | Prec@5%: \bm{$0.479 \pm 0.007$} -> 0.453 $\pm$ 0.007 | Hurt |
| fine-tune on 1000 small-world and 1000 original-prior examples (ablation) | DeepGraph-39 · Gaussian small-world graphs (n=35,p=39) | Prec@5%: \bm{$0.479 \pm 0.007$} -> \bm{$0.560 \pm 0.008$} | Helped |
| average predictions across input permutations (ablation) | DeepGraph-500 · Gaussian random graphs (n=50,p=500) | Prec@0.05%: 0.449 $\pm$ 0.018 -> $\bm{0.583 \pm 0.018}$ | Helped |
Closest papers by meaning
- Fast and Consistent Structure Learning in Graphical Models via Approximate Cross-Validation (NeurIPS 2026 · similarity 0.90)
- Graph Structure Inference with BAM: Neural Dependency Processing via Bilinear Attention (NeurIPS 2024 · similarity 0.90)
- Efficient Bayesian network structure learning via local Markov boundary search (NeurIPS 2021 · similarity 0.90)
- Distribution Learning Meets Graph Structure Sampling (NeurIPS 2025 · similarity 0.87)
- Guided Structural Inference: Leveraging Priors with Soft Gating Mechanisms (ICML 2025 · similarity 0.87)
- Learning Latent Graph Structures and their Uncertainty (ICML 2025 · 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.