Unified Representation of Molecules and Crystals for Machine Learning

Haoyan Huo, Matthias Rupp
4/21/2017
physics.chem-phcond-mat.mtrl-sci

Official code: hhaoyan/mbtr (14 stars, MIT, last commit 2022-11-15).

Abstract

Accurate simulations of atomistic systems from first principles are limited by computational cost. In high-throughput settings, machine learning can reduce these costs significantly by accurately interpolating between reference calculations. For this, kernel learning approaches crucially require a representation that accommodates arbitrary atomistic systems. We introduce a many-body tensor representation that is invariant to translations, rotations, and nuclear permutations of same elements, unique, differentiable, can represent molecules and crystals, and is fast to compute. Empirical evidence for competitive energy and force prediction errors is presented for changes in molecular structure, crystal chemistry, and molecular dynamics using kernel regression and symmetric gradient-domain machine learning as models. Applicability is demonstrated for phase diagrams of Pt-group/transition-metal binary systems.

View on arXivView PDF

Code Implementations(1)

hhaoyan/mbtrOfficial100%
MIT

A code for calculating MBTR molecule/crystal structure representation. (https://doi.org/10.1088/2632-2153/aca005)

140C++, TeXDec 27, 20183 years agoMIT
machine-learningmaterials-informaticsmaterials-sciencephysicspython+1 more

Similar Papers

Representations of molecules and materials for interpolation of quantum-mechanical simulations via machine learning

87% match1 repo332020

The dark side of the forces: assessing non-conservative force models for atomistic machine learning

85% match2024

Materials Learning Algorithms (MALA): Scalable Machine Learning for Electronic Structure Calculations in Large-Scale Atomistic Simulations

84% match2024

Where Should Physics Enter a Molecular Crystal Generator?

84% match1 repo12026

Forces are not Enough: Benchmark and Critical Evaluation for Machine Learning Force Fields with Molecular Simulations

84% match1 repo1162022

On the Interplay of Subset Selection and Informed Graph Neural Networks

84% match2023

Cite this paper

@article{huo2017unified,
  title  = {Unified Representation of Molecules and Crystals for Machine Learning},
  author = {Haoyan Huo and Matthias Rupp},
  year   = {2017},
  eprint = {1704.06439},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1704.06439}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1704.06439.svg)](https://codeofpaper.com/paper/1704.06439)

Discussion