Transfer learning of hyperparameters for fast construction of anisotropic GPR models: design and application to the machine-learned force field FFLUX,Physical Chemistry Chemical Physics

当前位置： X-MOL 学术 › Phys. Chem. Chem. Phys. › 论文详情

Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)

Transfer learning of hyperparameters for fast construction of anisotropic GPR models: design and application to the machine-learned force field FFLUX
Physical Chemistry Chemical Physics ( IF 2.9 ) Pub Date : 2024-09-03 , DOI: 10.1039/d4cp01862a
Bienfait K Isamura ₁ , Paul L A Popelier ₁

Affiliation

The polarisable machine-learned force field FFLUX requires pre-trained anisotropic Gaussian process regression (GPR) models of atomic energies and multipole moments to propagate unbiased molecular dynamics simulations. The outcome of FFLUX simulations is highly dependent on the predictive accuracy of the underlying models whose training entails determining the optimal set of model hyperparameters. Unfortunately, traditional direct learning (DL) procedures do not scale well on this task, especially when the hyperparameter search is initiated from a (set of) random guess solution(s). Additionally, the complexity of the hyperparameter space (HS) increases with the number of geometrical input features, at least for anisotropic kernels, making the optimization of hyperparameters even more challenging. In this study, we propose a transfer learning (TL) protocol that accelerates the training process of anisotropic GPR models by facilitating access to promising regions of the HS. The protocol is based on a seeding–relaxation mechanism in which an excellent guess solution is identified by rapidly building one or several small source models over a subset of the target training set before readjusting the previous guess over the entire set. We demonstrate the performance of this protocol by building and assessing the performance of DL and TL models of atomic energies and charges in various conformations of benzene, ethanol, formic acid dimer and the drug fomepizole. Our experiments suggest that TL models can be built one order of magnitude faster while preserving the quality of their DL analogs. Most importantly, when deployed in FFLUX simulations, TL models compete with or even outperform their DL analogs when it comes to performing FFLUX geometry optimization and computing harmonic vibrational modes.

中文翻译：

用于快速构建各向异性 GPR 模型的超参数迁移学习：机器学习力场 FFLUX 的设计和应用

可极化机器学习力场 FFLUX 需要预先训练的原子能和多极矩的各向异性高斯过程回归 (GPR) 模型来传播无偏分子动力学模拟。 FFLUX 模拟的结果高度依赖于基础模型的预测准确性，其训练需要确定模型超参数的最佳集。不幸的是，传统的直接学习（DL）程序在这项任务上不能很好地扩展，特别是当超参数搜索是从一组（一组）随机猜测解决方案启动时。此外，超参数空间（HS）的复杂性随着几何输入特征数量的增加而增加，至少对于各向异性核来说是这样，这使得超参数的优化更具挑战性。在这项研究中，我们提出了一种迁移学习（TL）协议，该协议通过促进对 HS 有希望的区域的访问来加速各向异性 GPR 模型的训练过程。该协议基于种子松弛机制，在该机制中，通过在目标训练集的子集上快速构建一个或多个小型源模型，然后重新调整对整个集合的先前猜测，来识别出色的猜测解决方案。我们通过构建和评估苯、乙醇、甲酸二聚体和药物甲吡唑各种构象的原子能和电荷的 DL 和 TL 模型的性能来展示该协议的性能。我们的实验表明，TL 模型的构建速度可以提高一个数量级，同时保持 DL 类似物的质量。最重要的是，当部署在 FFLUX 模拟中时，TL 模型在执行 FFLUX 几何优化和计算谐波振动模式方面可以与 DL 模拟竞争甚至优于它们。

更新日期：2024-09-04

点击分享查看原文

点击收藏

公开下载

阅读更多本刊新发论文本刊介绍/投稿指南