Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Fast Bayesian Inference of Sparse Networks with Automatic Sparsity Determination
Hang Yu, Songwei Wu, Luyin Xin, Justin Dauwels
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 |
|---|---|---|---|
| add BINCO stability selection to choose the penalty for G-ISTA | G-ISTA · Synthetic data (p=200, n=4p) | F1-score: 0.49 -> 0.72 | Improved |
| use BISN instead of G-ISTA+BINCO | BISN · Synthetic data (p=200, n=4p) | computational time (s): 5.73e2 -> 6.24 | Improved |
| use BISN instead of TIGER | BISN · Synthetic data (p=200, n=4p) | F1-score: 0.78 -> 0.96 | Improved |
| use BISN instead of G-ISTA+StARS at large dimension | BISN · Synthetic data (p=5000, n=4p) | computational time (s): 4.24e5 -> 9.36e3 | Improved |
| use BISN instead of G-ISTA+StARS (l1 shrinkage bias) | BISN · Synthetic data (p=5000, n=4p) | MSE of precision matrix: 4.34e-3 -> 1.04e-4 | Improved |
| reduce sample size from n=2p to n=p/4 for BISN | BISN · Synthetic data (p=1000) | F1-score: 0.99 -> 0.54 | Hurt |
| reduce sample size from n=2p to n=p/4 for BISN | BISN · Synthetic data (p=1000) | running time (s): 2.06e2 -> 4.79e2 | Hurt |
| reduce sample size from n=2p to n=p/4 for TIGER | TIGER · Synthetic data (p=1000) | running time (s): 3.13e5 -> 1.92e4 | Improved |
| use BISN instead of BINCO when n is small | BISN · Synthetic data (p=1000, n=p/4) | F1-score: 0.54 vs 0.56 | No clear effect |
| increase sample size tenfold to n=40p for G-ISTA with best penalty | G-ISTA · Synthetic data (p=1000) | F1-score: 0.99 | Mixed |
| use BISN instead of BINCO on stock network | BISN · S&P500 stocks crisis 2008-2009 | running time (s): 7.61e4 -> 3.68e2 | Improved |
Closest papers by meaning
- Beyond normality: Learning sparse probabilistic graphical models in the non-Gaussian setting (NeurIPS 2017 · similarity 0.90)
- Training Bayesian Neural Networks with Sparse Subspace Variational Inference (ICLR 2024 · similarity 0.89)
- Learning Identifiable Gaussian Bayesian Networks in Polynomial Time and Sample Complexity (NeurIPS 2017 · similarity 0.89)
- Graph Structure Inference with BAM: Neural Dependency Processing via Bilinear Attention (NeurIPS 2024 · similarity 0.88)
- Posterior Contraction Rates for sparse Kolmogorov-Arnold Networks in Anisotropic Besov Spaces (NeurIPS 2026 · similarity 0.87)
- Learning Local Neighborhoods of Non-Gaussian Graphical Models (AAAI 2025 · similarity 0.87)
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.
