Variational Entropy Search for Adjusting Expected Improvement

arXiv:2402.11345v1 Announce Type: cross
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 …

abstract acquisition adjusting arxiv bayesian box cs.lg domain entropy function functions improvement information math.oc max optimization search stat.ml type value