mlsanityData to 7 Oct 2026

Exploring Exploration in Bayesian Optimization

arXiv 2502.08208 ↗First posted 12 Feb 2025cs.LGcited by 8

Not on the map. Why?

Abstract (arXiv)

A well-balanced exploration-exploitation trade-off is crucial for successful acquisition functions in Bayesian optimization. However, there is a lack of quantitative measures for exploration, making it difficult to analyze and compare different acquisition functions. This work introduces two novel approaches - observation traveling salesman distance and observation entropy - to quantify the exploration characteristics of acquisition functions based on their selected observations. Using these measures, we examine the explorative nature of several well-known acquisition functions across a diverse set of black-box problems, uncover links between exploration and empirical performance, and reveal new relationships among existing acquisition functions. Beyond enabling a deeper understanding of acquisition functions, these measures also provide a foundation for guiding their design in a more principled and systematic manner.

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Builds on 24 of its 74 references that are in the map. First posted on arXiv 12 Feb 2025.

Builds on

24 of 74 references.

References · 24

  1. 2016The CMA Evolution Strategy: A Tutorialcited by 364
  2. 2019Scalable Global Optimization via Local Bayesian Optimizationcited by 359
  3. 2017Max-value Entropy Search for Efficient Bayesian Optimizationcited by 202
  4. 2014Predictive Entropy Search for Efficient Global Optimization of Black-box Functionscited by 201
  5. 2018Understanding the impact of entropy on policy optimizationcited by 171
  6. 2021High-Dimensional Bayesian Optimization with Sparse Axis-Aligned Subspacescited by 139
  7. 2017Batched Large-scale Bayesian Optimization in High-dimensional Spacescited by 116
  8. 2015Batch Bayesian Optimization via Local Penalizationcited by 103
  9. 2024Vanilla Bayesian Optimization Performs Great in High Dimensionscited by 87
  10. 2014Algorithms for multi-armed bandit problemscited by 63
  11. 2020Pathwise Conditioning of Gaussian Processescited by 63
  12. 2022Joint Entropy Search for Maximally-Informed Bayesian Optimizationcited by 55
12 more
  1. 2019Greed is Good: Exploration and Exploitation Trade-offs in Bayesian Optimisationcited by 47
  2. 2022Joint Entropy Search for Multi-objective Bayesian Optimizationcited by 45
  3. 2016Parallel Bayesian Global Optimization of Expensive Functionscited by 43
  4. 2020Efficient Nonmyopic Bayesian Optimization via One-Shot Multi-Step Treescited by 41
  5. 2021LassoBench: A High-Dimensional Hyperparameter Optimization Benchmark Suite for Lassocited by 41
  6. 2016Efficient multivariate entropy estimation via k-nearest neighbour distancescited by 26
  7. 2024Epsilon-Greedy Thompson Sampling to Bayesian Optimizationcited by 8
  8. 2016A sequential Monte Carlo approach to Thompson sampling for Bayesian optimizationcited by 7
  9. 2022PI is back! Switching Acquisition Functions in Bayesian Optimizationcited by 6
  10. 2023Computing Star Discrepancies with Numerical Black-Box Optimization Algorithmscited by 3
  11. 2025A Unified Framework for Entropy Search and Expected Improvement in Bayesian Optimizationcited by 3
  12. 2023On a Traveling Salesman Problem for Points in the Unit Cubecited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 4; 2026: 4.

8 papers.

  1. 2025Understanding High-Dimensional Bayesian Optimizationcited by 19
  2. 2025We Still Don't Understand High-Dimensional Bayesian Optimizationcited by 9
  3. 2025Dynamic Priors in Bayesian Optimization for Hyperparameter Optimizationcited by 2
  4. 2026Adaptive Candidate Point Thompson Sampling for High-Dimensional Bayesian Optimizationcited by 1
  5. 2025BOOST: A Data-Driven Framework for the Automated Joint Selection of Kernel and Acquisition Functions in Bayesian Optimizationcited by 0
  6. 2026Maximizing Reliability with Bayesian Optimizationcited by 0
  7. 2026Optimization of time-consuming experimental conditions using pseudo-experimental data guided by adaptive polynomial regressioncited by 0
  8. 2026Agentic High-Dimensional Bayesian Optimization with Hypothesis- and Evidence-Guided Searchcited by 0