arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1608.06194physics.chem-ph
收藏

理解机器学习中的分子表示:唯一性和目标相似性的作用

Understanding molecular representations in machine learning: The role of uniqueness and target similarity

Bing Huang, O. Anatole von Lilienfeld

摘要:

分子性质的机器学习(ML)模型的预测准确性取决于分子表示的选择。基于量子力学的假设,我们引入了一个层次的表示,满足唯一性和目标相似性标准。为了系统地控制目标相似性,我们依赖于原子间的多体扩展,如在通用力场中实施的,包括键合、角度和高阶项(BA)。高阶贡献的添加系统地增加了与真实势能的相似性和所得ML模型的预测准确性。我们报告了BAML模型性能的数值证据,这些模型是在电子相关和密度泛函理论水平上为数千个小有机分子预先计算的分子性质上训练的。研究的性质包括原子化的自由能、热容、零点振动能、偶极矩、极化率、HOMO/LUMO能量和能隙、电离势、电子亲和势和电子激发。经过训练后,BAML能够以前所未有的精度和速度预测样本外分子的能量或电子性质。

英文摘要:

The predictive accuracy of Machine Learning (ML) models of molecular properties depends on the choice of the molecular representation. Based on the postulates of quantum mechanics, we introduce a hierarchy of representations which meet uniqueness and target similarity criteria. To systematically control target similarity, we rely on interatomic many body expansions, as implemented in universal force-fields, including Bonding, Angular, and higher order terms (BA). Addition of higher order contributions systematically increases similarity to the true potential energy and predictive accuracy of the resulting ML models. We report numerical evidence for the performance of BAML models trained on molecular properties pre-calculated at electron-correlated and density functional theory level of theory for thousands of small organic molecules. Properties studied include enthalpies and free energies of atomization, heatcapacity, zero-point vibrational energies, dipole-moment, polarizability, HOMO/LUMO energies and gap, ionization potential, electron affinity, and electronic excitations. After training, BAML predicts energies or electronic properties of out-of-sample molecules with unprecedented accuracy and speed.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑