Research map / Probabilistic and scientific modeling / Bayesian optimization
Joint Entropy Search For Maximally-Informed Bayesian Optimization
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 |
|---|---|---|---|
| Consider the joint density over the optimum and its noiseless value rather than only one of them | Bayesian optimization · GP prior samples, 2D, 4D, 6D, and 12D | inference regret | Helped |
| Use joint entropy search instead of MES | Bayesian optimization · MLP tuning tasks | best observed classification accuracy | Helped |
| Use joint entropy search instead of EI | Bayesian optimization · MLP tuning tasks | best observed classification accuracy | Mixed |
| Use joint entropy search instead of MES | Bayesian optimization · Branin | simple regret | No clear effect |
| Use joint entropy search instead of competing methods | Bayesian optimization · Hartmann (3D) and Hartmann (6D) | simple regret | Helped |
| Use joint entropy search instead of PES | Bayesian optimization · GP sample tasks | runtime | Helped |
| Add occasional posterior-mean greedy selections (gamma-exploit) (ablation) | JES · Synthetic test functions | simple regret | Helped |
| Add a moderate fraction of posterior-mean greedy selections (gamma in {0.05, 0.1}) (ablation) | JES · Branin, Hartmann (3D), and Hartmann (6D) | inference regret | Helped |
| Increase exploitative selection fraction to gamma=0.2 (ablation) | JES · Hartmann (6D), Branin, and Hartmann (3D) | inference regret | Mixed |
| Add gamma-exploit selections (ablation) | JES · High-dimensional synthetic functions with model misspecification | simple regret | Helped |
| Vary the number of Monte Carlo optimum samples (ablation) | JES · GP sample tasks, especially higher-dimensional tasks | inference regret | Mixed |
| Use moment matching to approximate truncated-observation entropy rather than the MC approach (ablation) | JES · Entropy approximation analysis | entropy reduction | Hurt |
Closest papers by meaning
- Joint Entropy Search for Multi-Objective Bayesian Optimization (NeurIPS 2022 · similarity 0.94)
- Max-value Entropy Search for Efficient Bayesian Optimization (ICML 2017 · similarity 0.94)
- A Unified Framework for Entropy Search and Expected Improvement in Bayesian Optimization (ICML 2025 · similarity 0.93)
- Batched Energy-Entropy acquisition for Bayesian Optimization (NeurIPS 2024 · similarity 0.87)
- Cost-aware Bayesian Optimization via the Pandora's Box Gittins Index (NeurIPS 2024 · similarity 0.87)
- A General Framework for User-Guided Bayesian Optimization (ICLR 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.