Research map / Probabilistic and scientific modeling / Variational and generative sampling
Scalable Model Selection for Belief Networks
Zhao Song, Yusuke Muraoka, Ryohei Fujimaki, Lawrence Carin
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 FABIA model selection (gFIC shrinkage) to NVIL, 1-layer | NVIL · MNIST | test VLB (nats): -115.63 -> −114.96 | Improved |
| add FABIA model selection, 2-layer | NVIL · MNIST | test VLB (nats): -105.62 -> −104.92 | Improved |
| add FABIA model selection, 3-layer | NVIL · MNIST | test VLB (nats): -101.99 -> −101.14 | Improved |
| shrink model size via FABIA, 1-layer | NVIL · MNIST | test time (seconds): 8.47 -> 6.88 | Improved |
| shrink model size via FABIA, 2-layer | NVIL · MNIST | test time (seconds): 12.34 -> 9.18 | Improved |
| shrink model size via FABIA, 3-layer | NVIL · MNIST | test time (seconds): 15.66 -> 10.97 | Improved |
| use FABIA instead of VB with beta-normal prior shrinkage, 1-layer | VB · MNIST | test VLB (nats): -117.04 -> −114.96 | Improved |
| use FABIA instead of Rec-MCEM, 2-layer | Rec-MCEM · MNIST | test VLB (nats): -106.54 -> −104.92 | Improved |
| initialize NVIL with true model size instead of FABIA's initial size | NVIL · synthetic one- and two-layer SBN | test log-likelihood | Improved |
| FABIA vs NVIL (True) with the true model size | FABIA · synthetic one- and two-layer SBN | test log-likelihood convergence | No clear effect |
| use FABIA instead of NVIL | NVIL · RCV1 | test perplexity: 857 -> 856 | No clear effect |
| use FABIA instead of NVIL | NVIL · Wikipedia | test perplexity: 735 -> 730 | No clear effect |
Closest papers by meaning
- An Asymptotically Optimal Coordinate Descent Algorithm for Learning Bayesian Networks from Gaussian Models (JMLR 2025 · similarity 0.85)
- Optimizing Regularized Cholesky Score for Order-Based Learning of Bayesian Networks (TPAMI 2020 · similarity 0.85)
- Learning Identifiable Gaussian Bayesian Networks in Polynomial Time and Sample Complexity (NeurIPS 2017 · similarity 0.85)
- Finding All Bayesian Network Structures within a Factor of Optimal (AAAI 2019 · similarity 0.85)
- Learning Fast-Inference Bayesian Networks (NeurIPS 2021 · similarity 0.85)
- A Study of Bayesian Neural Network Surrogates for Bayesian Optimization (ICLR 2024 · similarity 0.84)
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.
