Research map / Probabilistic and scientific modeling / Bayesian optimization
Maximizing acquisition functions for Bayesian optimization
James Wilson, Frank Hutter, Marc Deisenroth
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 |
|---|---|---|---|
| Use gradient-based stochastic optimization (Adam) instead of gradient-free or random-search maximization | Parallel Monte Carlo EI on synthetic and black-box tasks · Synthetic and black-box tasks | BO performance | Improved |
| Use greedy parallel maximization instead of joint maximization | Myopic maximal acquisition functions · Synthetic and black-box tasks | BO performance | Improved |
| Increase acquisition dimensionality | Gradient-based and submodular acquisition maximizers · Synthetic and black-box tasks | BO performance gain | Improved |
| Combine gradient-based optimization with greedy submodular maximization | Parallel acquisition maximization · Lower-dimensional cases | BO performance | No clear effect |
| Use parallel UCB instead of parallel EI | Bayesian optimization · All tasks except Levy | BO performance | Improved |
| Use parallel UCB instead of parallel EI | Bayesian optimization · Levy | BO performance | Mixed |
| Use incremental rather than joint parallel EI | Greedy parallel EI · Higher-dimensional acquisition surfaces | BO performance | Improved |
| Initialize query sets by sampling from the marginal acquisition (ablation) | Query optimizers · Experiments across tasks | BO performance | Improved |
| Fantasize pending-query outcomes at their predictive mean (ablation) | Greedy parallel acquisition maximization · Experiments across tasks | BO performance | No clear effect |
| Use stochastic rather than deterministic minibatches (ablation) | Adam and CMA-ES acquisition maximization · Acquisition optimization experiments | Acquisition maximization performance | Improved |
| Use Adam rather than L-BFGS-B or Polyak averaging (ablation) | Gradient-based acquisition maximization · Experiments across tasks | Acquisition maximization performance | Improved |
| Use CMA-ES rather than DIRECT (ablation) | Black-box acquisition maximization · Experiments across tasks | Acquisition maximization performance | Improved |
Closest papers by meaning
- FunBO: Discovering Acquisition Functions for Bayesian Optimization with FunSearch (ICML 2025 · similarity 0.91)
- A Unified Framework for Entropy Search and Expected Improvement in Bayesian Optimization (ICML 2025 · similarity 0.91)
- Simple Extensions of Single-Objective Acquisition Functions and Hedge Strategies for Multi-Objective Bayesian Optimization (NeurIPS 2026 · similarity 0.90)
- Batched Energy-Entropy acquisition for Bayesian Optimization (NeurIPS 2024 · similarity 0.90)
- Random Exploration in Bayesian Optimization: Order-Optimal Regret and Computational Efficiency (ICML 2024 · similarity 0.88)
- Bayesian Optimization of Function Networks with Partial Evaluations (ICML 2024 · similarity 0.88)
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.
