Diagnosing and fixing common problems in Bayesian optimization for molecule design

Austin Tripp, José Miguel Hernández-Lobato

Keywords: Bayesian optimization, molecule, chemistry

TL;DR: Basic Bayesian optimization works better than people think, *if* hyperparameters are tuned correctly

Abstract: 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.

Submission Number: 144