Research map / Probabilistic and scientific modeling / Bayesian optimization
Efficient Nonmyopic Bayesian Optimization via One-Shot Multi-Step Trees
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 |
|---|---|---|---|
| Use multi-step lookahead instead of one-step EI | Bayesian optimization · Synthetic functions (average) | GAP: 0.576 -> 0.753 | Helped |
| Increase lookahead from 2-step to 3-step (ablation) | One-shot multi-step Bayesian optimization · Synthetic functions (average) | GAP: 0.725 -> 0.747 | Helped |
| Increase lookahead from 3-step to 4-step (ablation) | One-shot multi-step Bayesian optimization · Synthetic functions (average) | GAP: 0.747 -> 0.753 | No clear effect |
| Use one fantasy sample per stage rather than multiple samples (ablation) | Multi-step path · Synthetic functions (average) | GAP: 0.747 -> 0.761 | Helped |
| Use one fantasy sample per stage rather than multiple samples (ablation) | Multi-step path · Synthetic functions (average) | time per iteration (seconds): 197.7 -> 17.50 | Helped |
| Approximate later adaptive decisions with a non-adaptive q-EI batch (ablation) | ENO · Synthetic functions (average) | GAP: 0.576 -> 0.763 | Helped |
| Approximate later adaptive decisions with a non-adaptive q-EI batch (ablation) | ENO · Synthetic functions (average) | time per iteration (seconds): 197.7 -> 15.61 | Helped |
| Use a multi-step path instead of one-step EI | Bayesian optimization · Real functions with EI GAP < 0.8 (average) | GAP: 0.448 -> 0.523 | Helped |
Closest papers by meaning
- BayeSQP: Bayesian Optimization through Sequential Quadratic Programming (NeurIPS 2025 · similarity 0.90)
- BARK: A Fully Bayesian Tree Kernel for Black-box Optimization (ICML 2025 · similarity 0.88)
- A General Framework for User-Guided Bayesian Optimization (ICLR 2024 · similarity 0.87)
- BONSAI: Bayesian Optimization with Natural Simplicity and Interpretability (NeurIPS 2026 · similarity 0.87)
- BOFormer: Learning to Solve Multi-Objective Bayesian Optimization via Non-Markovian RL (ICLR 2025 · similarity 0.87)
- Cost-aware Bayesian Optimization via the Pandora's Box Gittins Index (NeurIPS 2024 · 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.