Bayesian optimization (BO) is a principled approach to molecular design tasks. In this paper we explain three pitfalls of BO which can cause poor empirical performance: an incorrect prior width, over-smoothing, and inadequate acquisition function maximization. We show that with these issues addressed, even a basic BO setup is able to achieve the highest overall performance on the PMO benchmark for molecule design (Gao et al 2022). These results suggest that BO may benefit from more attention in the machine learning for molecules community.

翻译：贝叶斯优化（BO）是分子设计任务中一种基于原理的方法。本文阐释了导致BO实证性能不佳的三个常见缺陷：先验宽度设置不当、过度平滑现象以及采集函数最大化不足。研究表明，在解决这些问题后，即使采用基础BO配置，也能在分子设计PMO基准测试（Gao等人，2022）中取得最佳综合性能。这些结果表明，BO方法值得分子机器学习领域给予更多关注。