A study of EHVI vs fixed scalarization for molecule design

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yong, Anabel, Tripp, Austin, Hosseini-Gerami, Layla, Paige, Brooks
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917168019406848
author Yong, Anabel
Tripp, Austin
Hosseini-Gerami, Layla
Paige, Brooks
author_facet Yong, Anabel
Tripp, Austin
Hosseini-Gerami, Layla
Paige, Brooks
contents Multi-objective Bayesian optimization (MOBO) provides a principled framework for navigating trade-offs in molecular design. However, its empirical advantages over scalarized alternatives remain underexplored. We benchmark a simple Pareto-based MOBO strategy - Expected Hypervolume Improvement (EHVI) - against a simple fixed-weight scalarized baseline using Expected Improvement (EI), under a tightly controlled setup with identical Gaussian Process surrogates and molecular representations. Across three molecular optimization tasks, EHVI consistently outperforms scalarized EI in terms of Pareto front coverage, convergence speed, and chemical diversity. While scalarization encompasses flexible variants - including random or adaptive schemes - our results show that even strong deterministic instantiations can underperform in low-data regimes. These findings offer concrete evidence for the practical advantages of Pareto-aware acquisition in de novo molecular optimization, especially when evaluation budgets are limited and trade-offs are nontrivial.
format Preprint
id arxiv_https___arxiv_org_abs_2507_13704
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A study of EHVI vs fixed scalarization for molecule design
Yong, Anabel
Tripp, Austin
Hosseini-Gerami, Layla
Paige, Brooks
Machine Learning
Multi-objective Bayesian optimization (MOBO) provides a principled framework for navigating trade-offs in molecular design. However, its empirical advantages over scalarized alternatives remain underexplored. We benchmark a simple Pareto-based MOBO strategy - Expected Hypervolume Improvement (EHVI) - against a simple fixed-weight scalarized baseline using Expected Improvement (EI), under a tightly controlled setup with identical Gaussian Process surrogates and molecular representations. Across three molecular optimization tasks, EHVI consistently outperforms scalarized EI in terms of Pareto front coverage, convergence speed, and chemical diversity. While scalarization encompasses flexible variants - including random or adaptive schemes - our results show that even strong deterministic instantiations can underperform in low-data regimes. These findings offer concrete evidence for the practical advantages of Pareto-aware acquisition in de novo molecular optimization, especially when evaluation budgets are limited and trade-offs are nontrivial.
title A study of EHVI vs fixed scalarization for molecule design
topic Machine Learning
url https://arxiv.org/abs/2507.13704