Research map / Probabilistic and scientific modeling / Bayesian optimization
Cost-aware Stopping for Bayesian Optimization
Qian Xie, Linda Cai, Alexander Terenin, Peter I. Frazier, Ziv Scully
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 |
|---|---|---|---|
| check stopping after the posterior update rather than before (ablation) | PBGI stopping rule · 1D Bayesian regret (single GP draw) | cost-adjusted regret | Improved |
| pair stopping rule with PBGI instead of LogEIPC acquisition | PBGI/LogEIPC stopping rule · LCBench / NATS-Bench | cost-adjusted simple regret | Improved |
| use PBGI/LogEIPC stopping rule instead of baseline stopping rules | PBGI acquisition · 8D Bayesian regret | cost-adjusted regret | Improved |
| use PBGI/LogEIPC stopping rule instead of baselines at high cost scale | PBGI/LogEIPC acquisition · 1D Bayesian regret | cost-adjusted regret | Improved |
| use PBGI/LogEIPC stopping rule on small LCBench datasets | PBGI/LogEIPC stopping rule · LCBench (datasets <10000 instances) | cost-adjusted regret | Hurt |
| use PBGI/LogEIPC stopping rule on two NATS datasets | PBGI acquisition · NATS-Bench (two non-cifar10 datasets) | cost-adjusted regret | Hurt |
| use PBGI/LogEIPC stopping rule instead of SRGap-med and UCB-LCB | PBGI acquisition · NATS-Bench (200-iteration cap) | stopping before cap | Improved |
| use PBGI/LogEIPC stopping rule instead of PRB | stopping rule computation · 8D Bayesian regret | CPU time per iteration | Improved |
| pair with PBGI vs LogEIPC at lambda=1e-5 | PBGI/LogEIPC stopping rule · LCBench (large datasets, Top-1 share) | % tasks ranked Top-1: 20.0 -> 50.0 | Improved |
Closest papers by meaning
- Cost-aware Bayesian Optimization via the Pandora's Box Gittins Index (NeurIPS 2024 · similarity 0.90)
- Stopping Bayesian Optimization with Probabilistic Regret Bounds (NeurIPS 2024 · similarity 0.90)
- BONSAI: Bayesian Optimization with Natural Simplicity and Interpretability (NeurIPS 2026 · similarity 0.90)
- Bayesian Optimization for Unknown Cost-Varying Variable Subsets with No-Regret Costs (AAAI 2025 · similarity 0.89)
- Bayesian Optimisation with Unknown Hyperparameters: Regret Bounds Logarithmically Closer to Optimal (NeurIPS 2024 · similarity 0.88)
- A Unified Framework for Entropy Search and Expected Improvement in Bayesian Optimization (ICML 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.
