将专家经验与元贝叶斯优化结合，加速融合能源等高成本科学领域的实验发现。
Human-in-the-Loop Meta Bayesian Optimization for Fusion Energy and Scientific Applications
- 用元学习构建可复用的代理模型，融合专家先验指导实验设计。
- 在惯性约束聚变中提升能量输出，优于现有贝叶斯优化方法。
- 提供可解释建议，适合需要高可靠性、数据稀缺的科学实验场景。
惯性约束聚变（ICF）有望实现可持续、近乎无限的清洁能源，但受限于高昂成本和有限的实验机会。本文提出人类在环路的元贝叶斯优化（HL-MBO），将专家知识与少样本、不确定性感知的机器学习相结合，加速数据稀缺、高风险科学领域中的发现进程。HL-MBO引入元学习的代理模型与专家驱动的采集函数，以推荐候选实验；为增强信任并支持决策，还提供建议的可解释性说明。实验表明，该方法在ICF能量产率优化上超越现有贝叶斯优化技术，并在分子优化和超导材料临界温度最大化基准任务中表现优异。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。