Research map / Probabilistic and scientific modeling / Bayesian optimization
Bayesian Optimization with High-Dimensional Outputs
Wesley Maddox, Maximilian Balandat, Andrew Gordon Wilson, Eytan Bakshy
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 |
|---|---|---|---|
| Model correlated objectives and constraints jointly with an MTGP instead of batch-independent GPs | Constrained multi-objective BO · C2DTLZ2, q=2 | hypervolume | Improved |
| Model correlated objectives and constraints jointly with an MTGP instead of batch-independent GPs | Constrained multi-objective BO · OSY, q=10 | hypervolume | Improved |
| Replace independent GPs with a joint MTGP | SCBO · Lunar Lander, m=50 | steps to at least 50% chance of feasibility: about 75 steps | Improved |
| Replace independent GPs with a joint MTGP | SCBO · Lunar Lander, m=100 | feasibility: 350 steps | Improved |
| Replace independent GPs with a joint MTGP | SCBO · Lunar Lander, m=50 | best merit | No clear effect |
| Replace independent GPs with a joint MTGP | SCBO · MOPTA08 | time to feasibility | Improved |
| Replace independent GPs with a joint MTGP | SCBO · MOPTA08 | best objective | Improved |
| Optimize constituent high-dimensional outputs with a composite HOGP surrogate instead of optimizing the scalar metric directly | Composite BO · Chemical Pollutants | mean squared error to target output grid | Improved |
| Optimize constituent high-dimensional outputs with a composite HOGP surrogate instead of optimizing the scalar metric directly | Composite BO · PDE control | weighted variance of PDE output | Improved |
| Optimize constituent high-dimensional outputs with a composite HOGP surrogate instead of optimizing the scalar metric directly | Composite BO · Cell-tower coverage | coverage quality | Improved |
| Add a GP prior encouraging smoothly varying latent dimensions (ablation) | HOGP-CF · Chemical Pollutants, PDE control, and cell-tower coverage | optimization performance | Improved |
| Add a GP prior encouraging smoothly varying latent dimensions (ablation) | HOGP-CF · Optical Interferometer | optimization performance | Hurt |
Closest papers by meaning
- Understanding High-Dimensional Bayesian Optimization (ICML 2025 · similarity 0.90)
- Scalable Bayesian Optimization via Focalized Sparse Gaussian Processes (NeurIPS 2024 · similarity 0.90)
- Optimizing Posterior Samples for Bayesian Optimization via Rootfinding (ICLR 2025 · similarity 0.88)
- Sample-efficient Bayesian Optimisation Using Known Invariances (NeurIPS 2024 · similarity 0.88)
- Bayesian Optimization via Continual Variational Last Layer Training (ICLR 2025 · similarity 0.88)
- Standard Gaussian Process is All You Need for High-Dimensional Bayesian Optimization (ICLR 2025 · 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.
