English

Variational Entropy Search for Adjusting Expected Improvement

Machine Learning 2025-03-11 v2 Machine Learning Optimization and Control

Abstract

Bayesian optimization is a widely used technique for optimizing black-box functions, with Expected Improvement (EI) being the most commonly utilized acquisition function in this domain. While EI is often viewed as distinct from other information-theoretic acquisition functions, such as entropy search (ES) and max-value entropy search (MES), our work reveals that EI can be considered a special case of MES when approached through variational inference (VI). In this context, we have developed the Variational Entropy Search (VES) methodology and the VES-Gamma algorithm, which adapts EI by incorporating principles from information-theoretic concepts. The efficacy of VES-Gamma is demonstrated across a variety of test functions and read datasets, highlighting its theoretical and practical utilities in Bayesian optimization scenarios.

Keywords

Cite

@article{arxiv.2402.11345,
  title  = {Variational Entropy Search for Adjusting Expected Improvement},
  author = {Nuojin Cheng and Stephen Becker},
  journal= {arXiv preprint arXiv:2402.11345},
  year   = {2025}
}

Comments

This is a preliminary technical report. For a more comprehensive study, please refer to arXiv:2501.18756

R2 v1 2026-06-28T14:51:53.702Z