Research map / Probabilistic and scientific modeling / Bayesian optimization
Bayesian optimization for automated model selection
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 |
|---|---|---|---|
| Predict model evidence with a GP using the Hellinger-distance kernel instead of mean prediction | GP over model space · CONCRETE 20% train | RMSE: 0.109 -> 0.107 | Helped |
| Same change at 80% training models | GP over model space · HOUSING 80% train | RMSE: 0.206 -> 0.114 | Helped |
| Same change on MAUNA LOA at 80% training models | GP over model space · MAUNA LOA 80% train | RMSE: 0.534 -> 0.474 | Helped |
| Use k-NN regression with Hellinger distance instead of mean prediction | k-NN (Hellinger) · HOUSING 20% train | RMSE: 0.210 -> 0.347 | Hurt |
| Use k-NN with shortest-path distance instead of mean prediction | k-NN (SP) · MAUNA LOA 80% train | RMSE: 0.534 -> 1.207 | Hurt |
| Increase fraction of models used for training from 20% to 80% | GP (Hellinger) · HOUSING | RMSE: 0.175 -> 0.114 | Helped |
| Increase training fraction from 20% to 80% for k-NN (Hellinger) | k-NN (Hellinger) · CONCRETE | RMSE: 0.233 -> 0.200 | Helped |
| Increase training fraction from 20% to 80% for k-NN (SP) | k-NN (SP) · MAUNA LOA | RMSE: 0.736 -> 1.207 | Hurt |
| Use Bayesian optimization over model space (BOMS) instead of greedy compositional kernel search (CKS) | BOMS · six regression datasets, 50 evaluations | best model evidence | Helped |
| Use BOMS instead of CKS on MAUNA LOA | BOMS · MAUNA LOA | best model evidence | No clear effect |
| Use BOMS instead of CKS on METHANE | BOMS · METHANE | best model evidence | Mixed |
Closest papers by meaning
- Structural Kernel Search via Bayesian Optimization and Symbolical Optimal Transport (NeurIPS 2022 · similarity 0.91)
- Kernel-based guarantees for nonlinear parametric models in Bayesian optimization (NeurIPS 2026 · similarity 0.89)
- Automated Kernel Discovery Towards Understanding High-dimensional Bayesian Optimization (NeurIPS 2026 · similarity 0.89)
- Data Mixture Optimization: A Multi-fidelity Multi-scale Bayesian Framework (NeurIPS 2025 · similarity 0.87)
- Adaptive Kernel Design for Bayesian Optimization Is a Piece of CAKE with LLMs (NeurIPS 2025 · similarity 0.86)
- Bivariate Causal Discovery using Bayesian Model Selection (ICML 2024 · 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.