Human-in-the-Loop Meta Bayesian Optimization for Fusion Energy and Scientific Applications

TL;DR

This paper introduces HL-MBO, a human-in-the-loop Bayesian optimization framework that combines expert knowledge and machine learning to accelerate scientific discovery in energy and materials research.

Contribution

It presents a novel meta-learned surrogate model with an expert-informed acquisition function and interpretable explanations, improving optimization in data-scarce scientific domains.

Findings

HL-MBO outperforms existing Bayesian optimization methods in ICF energy yield.

HL-MBO demonstrates superior performance on molecular optimization benchmarks.

HL-MBO effectively maximizes critical temperature in superconducting materials.

Abstract

Inertial Confinement Fusion (ICF) holds transformative promise for sustainable, near-limitless clean energy, yet remains constrained by prohibitively high costs and limited experimental opportunities. This paper presents Human-in-the-Loop Meta Bayesian Optimization (HL-MBO), a framework that integrates expert knowledge with few-shot, uncertainty-aware machine learning to accelerate discovery in data-scarce, high-stakes scientific domains. HL-MBO introduces a meta-learned surrogate model with an expert-informed acquisition function to recommend candidate experiments. To foster trust and enable informed decisions, HL-MBO also provides interpretable explanations of its suggestions. We show HL-MBO outperforms current BO methods on ICF energy yield optimization, as well as benchmarks in molecular optimization and critical temperature maximization for superconducting materials.