Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Bayesian Joint Matrix Decomposition for Data Integration with Heterogeneous Noise
Chihao Zhang, Shihua Zhang
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 |
|---|---|---|---|
| model per-source noise variance instead of concatenating sources (ablation) | VI-BJMD · Small synthetic source 3 | AUC: 72.79 -> 79.16 | Improved |
| model per-source noise variance instead of concatenating sources (ablation) | MAP-BJMD · Small synthetic source 3 | AUC: 69.79 -> 77.85 | Improved |
| model per-source noise variance instead of concatenating sources (ablation) | VI-BJMD · Large synthetic source 3 | AUC: 77.97 -> 91.85 | Improved |
| model per-source noise variance instead of concatenating sources (ablation) | MAP-BJMD · Large synthetic source 3 | AUC: 72.45 -> 89.09 | Improved |
| model per-source noise variance instead of concatenating sources (ablation) | MAP-BJMD · Small synthetic source 1 | AUC: 92.04 -> 99.67 | Improved |
| model per-source noise variance (JMD vs GLAD) (ablation) | GLAD · Small synthetic source 3 | AUC: 66.35 -> 72.06 | Improved |
| treat noise variance as random variable (full Bayesian) rather than hyperparameter | JMD · Small synthetic source 3 | AUC: 72.06 -> 79.16 | Improved |
| use MAP instead of variational inference | VI-BJMD · Small synthetic | time (s): 230.21 -> 4.20 | Improved |
| use VI instead of MAP on large synthetic data | MAP-BJMD · Large synthetic source 3 | AUC: 89.09 -> 91.85 | Improved |
| use VI instead of MAP on real data | MAP-BJMD · 3Sources Guardian and Yale side | AUC (%): 88.82 -> 90.78 | Mixed |
| noise-guided feature selection (19466 to 4210 genes) | MAP-BJMD · METABRIC low cellularity | AUC (%): 82.22 -> 84.34 | Improved |
| noise-guided feature selection (19466 to 4210 genes) | Joint-NMF · METABRIC low cellularity | AUC (%): 65.81 -> 82.70 | Improved |
Closest papers by meaning
- Partial Information Decomposition via Normalizing Flows in Latent Gaussian Distributions (NeurIPS 2025 · similarity 0.86)
- Functional Bayesian Tucker Decomposition for Continuous-indexed Tensor Data (ICLR 2024 · similarity 0.85)
- Variational Inference with Mixtures of Isotropic Gaussians (NeurIPS 2025 · similarity 0.84)
- Heterogeneous Sufficient Dimension Reduction and Subspace Clustering (ICML 2025 · similarity 0.84)
- Covariate-moderated Empirical Bayes Matrix Factorization (NeurIPS 2025 · similarity 0.84)
- Data Mixture Optimization: A Multi-fidelity Multi-scale Bayesian Framework (NeurIPS 2025 · 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.
