Cross-platform hyperparameter optimization for machine learning interatomic potentials.
Cross-platform hyperparameter optimization for machine learning interatomic potentials.
复制标题
DOI：
10.1063/5.0155618
复制
发表时间：
2023-07
期刊：
影响因子：
--
通讯作者：
Daniel F Thomas du Toit;Volker L. Deringer
中科院分区：
文献类型：
--
作者：
Daniel F Thomas du Toit;Volker L. Deringer
基于机器学习 (ML) 的原子间势在材料建模中越来越受欢迎，可以对成千上万个原子进行高精度模拟。然而，机器学习势的性能在很大程度上取决于超参数的选择，即模型遇到数据之前设置的参数。当超参数没有直观的物理解释并且相应的优化空间很大时，这个问题尤其严重。在这里，我们描述了一个公开可用的 Python 包，该包有助于跨不同 ML 潜在拟合框架进行超参数优化。我们讨论与优化本身和验证数据的选择相关的方法学方面，并展示示例应用程序。我们预计该软件包将成为更广泛的计算框架的一部分，以加速机器学习潜力在物理科学中的主流适应。
Machine-learning (ML)-based interatomic potentials are increasingly popular in material modeling, enabling highly accurate simulations with thousands and millions of atoms. However, the performance of machine-learned potentials depends strongly on the choice of hyperparameters-that is, of those parameters that are set before the model encounters data. This problem is particularly acute where hyperparameters have no intuitive physical interpretation and where the corresponding optimization space is large. Here, we describe an openly available Python package that facilitates hyperparameter optimization across different ML potential fitting frameworks. We discuss methodological aspects relating to the optimization itself and to the selection of validation data, and we show example applications. We expect this package to become part of a wider computational framework to speed up the mainstream adaptation of ML potentials in the physical sciences.