Research map / Probabilistic and scientific modeling / Bayesian optimization
Gaussian Process Planning with Lipschitz Continuous Reward Functions: Towards Unifying Bayesian Optimization, Active Learning, and Beyond
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 adaptive ε-GPP planning rather than maximum-likelihood observations during planning | Gaussian-process planning · Wind speed field | total rewards | Helped |
| Use a longer online planning horizon instead of greedy planning with H'=1 (ablation) | ε-GPP · Wind speed field | total rewards | Helped |
| Use adaptive ε-GPP planning rather than maximum-likelihood observations during planning | Gaussian-process planning · lg-K field | total normalized rewards | Helped |
| Use the anytime ε-GPP variant | Gaussian-process planning · lg-K field | total normalized rewards | Helped |
| Increase the planning horizon beyond greedy H'=1 (ablation) | ε-GPP with β=0 · lg-K field | total normalized rewards | Helped |
| Use a moderate exploration weight β=2 instead of β=0 (ablation) | ε-GPP · lg-K field | total normalized rewards | Helped |
| Use aggressive exploration with β≥10 (ablation) | ε-GPP · lg-K field | total normalized rewards | Hurt |
| Tighten the loss bound ε (ablation) | ε-GPP · Wind speed field | total rewards | Helped |
| Tighten ε while holding planning horizon H'=2 (ablation) | ε-GPP · Wind speed field | total rewards | Mixed |
| Use greedy planning with H'=1 and β=0 (ablation) | ε-GPP · lg-K field | total normalized rewards | Hurt |
Closest papers by meaning
- Transition Constrained Bayesian Optimization via Markov Decision Processes (NeurIPS 2024 · similarity 0.88)
- Optimizing the Unknown: Black Box Bayesian Optimization with Energy-Based Model and Reinforcement Learning (NeurIPS 2025 · similarity 0.87)
- Sequential Monte Carlo for Policy Optimization in Continuous POMDPs (NeurIPS 2025 · similarity 0.86)
- Bayesian Analysis of Combinatorial Gaussian Process Bandits (ICLR 2025 · similarity 0.86)
- BayeSQP: Bayesian Optimization through Sequential Quadratic Programming (NeurIPS 2025 · similarity 0.85)
- Improved Regret Analysis in Gaussian Process Bandits: Optimality for Noiseless Reward, RKHS norm, and Non-Stationary Variance (ICML 2025 · similarity 0.85)
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.