Fast,accurate,and transferable many-body interatomic potentials by symbolic regression

作     者：Alberto Hernandez Adarsh Balasubramanian Fenglin Yuan Simon A.M.Mason Tim Mueller 

作者机构：Department of Materials Science and EngineeringJohns Hopkins UniversityBaltimoreMDUSA 

出 版 物：《npj Computational Materials》 (计算材料学（英文）)

年 卷 期：2019年第5卷第1期

页      面：161-171页

核心收录：

学科分类：0711[理学-系统科学] 07[理学] 08[工学] 0805[工学-材料科学与工程（可授工学、理学学位）] 081101[工学-控制理论与控制工程] 0701[理学-数学] 0811[工学-控制科学与工程] 0801[工学-力学（可授工学、理学学位）] 0702[理学-物理学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 071102[理学-系统分析与集成] 081103[工学-系统工程] 

基　　金：We acknowledge financial support from the Office of Naval Research grant number N000141512665 

主　　题：symbolic simplicity transfer 

摘      要：The length and time scales of atomistic simulations are limited by the computational cost of the methods used to predict material *** recent years there has been great progress in the use of machine-learning algorithms to develop fast and accurate interatomic potential models,but it remains a challenge to develop models that generalize well and are fast enough to be used at extreme time and length *** address this challenge,we have developed a machine-learning algorithm based on symbolic regression in the form of genetic programming that is capable of discovering accurate,computationally efficient many-body potential *** key to our approach is to explore a hypothesis space of models based on fundamental physical principles and select models within this hypothesis space based on their accuracy,speed,and *** focus on simplicity reduces the risk of overfitting the training data and increases the chances of discovering a model that generalizes *** algorithm was validated by rediscovering an exact Lennard-Jones potential and a Sutton-Chen embedded-atom method potential from training data generated using these *** using training data generated from density functional theory calculations,we found potential models for elemental copper that are simple,as fast as embedded-atom models,and capable of accurately predicting properties outside of their training *** approach requires relatively small sets of training data,making it possible to generate training data using highly accurate methods at a reasonable computational *** present our approach,the forms of the discovered models,and assessments of their transferability,accuracy and speed.