Diagnosing and fixing common problems in Bayesian optimization for molecule design
This work improves Bayesian optimization for molecule design, which is an incremental advancement for researchers in machine learning and chemistry.
The paper identified three common pitfalls in Bayesian optimization for molecule design and demonstrated that addressing them enables a basic BO setup to achieve the highest overall performance on the PMO benchmark.
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.