Active learning of uniformly accurate interatomic potentials for materials simulation

Linfeng Zhang, De Ye Lin, Han Wang, Roberto Car, Weinan E

Research output: Contribution to journalArticle

Abstract

An active learning procedure called deep potential generator (DP-GEN) is proposed for the construction of accurate and transferable machine learning-based models of the potential energy surface (PES) for the molecular modeling of materials. This procedure consists of three main components: Exploration, generation of accurate reference data, and training. Application to the sample systems of Al, Mg, and Al-Mg alloys demonstrates that DP-GEN can produce uniformly accurate PES models with a minimal number of reference data.

Original languageEnglish (US)
Article number023804
JournalPhysical Review Materials
Volume3
Issue number2
DOIs
StatePublished - Feb 25 2019

Fingerprint

Potential energy surfaces
Molecular modeling
Learning systems
Problem-Based Learning

Cite this

@article{360318c745ed40b7a83db3f71ca70208,
title = "Active learning of uniformly accurate interatomic potentials for materials simulation",
abstract = "An active learning procedure called deep potential generator (DP-GEN) is proposed for the construction of accurate and transferable machine learning-based models of the potential energy surface (PES) for the molecular modeling of materials. This procedure consists of three main components: Exploration, generation of accurate reference data, and training. Application to the sample systems of Al, Mg, and Al-Mg alloys demonstrates that DP-GEN can produce uniformly accurate PES models with a minimal number of reference data.",
author = "Linfeng Zhang and Lin, {De Ye} and Han Wang and Roberto Car and Weinan E",
year = "2019",
month = "2",
day = "25",
doi = "https://doi.org/10.1103/PhysRevMaterials.3.023804",
language = "English (US)",
volume = "3",
journal = "Physical Review Materials",
issn = "2475-9953",
number = "2",

}

Active learning of uniformly accurate interatomic potentials for materials simulation. / Zhang, Linfeng; Lin, De Ye; Wang, Han; Car, Roberto; E, Weinan.

In: Physical Review Materials, Vol. 3, No. 2, 023804, 25.02.2019.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Active learning of uniformly accurate interatomic potentials for materials simulation

AU - Zhang, Linfeng

AU - Lin, De Ye

AU - Wang, Han

AU - Car, Roberto

AU - E, Weinan

PY - 2019/2/25

Y1 - 2019/2/25

N2 - An active learning procedure called deep potential generator (DP-GEN) is proposed for the construction of accurate and transferable machine learning-based models of the potential energy surface (PES) for the molecular modeling of materials. This procedure consists of three main components: Exploration, generation of accurate reference data, and training. Application to the sample systems of Al, Mg, and Al-Mg alloys demonstrates that DP-GEN can produce uniformly accurate PES models with a minimal number of reference data.

AB - An active learning procedure called deep potential generator (DP-GEN) is proposed for the construction of accurate and transferable machine learning-based models of the potential energy surface (PES) for the molecular modeling of materials. This procedure consists of three main components: Exploration, generation of accurate reference data, and training. Application to the sample systems of Al, Mg, and Al-Mg alloys demonstrates that DP-GEN can produce uniformly accurate PES models with a minimal number of reference data.

UR - http://www.scopus.com/inward/record.url?scp=85062498648&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85062498648&partnerID=8YFLogxK

U2 - https://doi.org/10.1103/PhysRevMaterials.3.023804

DO - https://doi.org/10.1103/PhysRevMaterials.3.023804

M3 - Article

VL - 3

JO - Physical Review Materials

JF - Physical Review Materials

SN - 2475-9953

IS - 2

M1 - 023804

ER -