Strategies for the Construction of Machine-Learning Potentials for Accurate and Efficient Atomic-Scale Simulations

Abstract

Recent advances in machine-learning interatomic potentials have enabled the efficient modeling of complex atomistic systems with an accuracy that is comparable to that of conventional quantum mechanics based methods. At the same time, the construction of new machine-learning potentials can seem a daunting task, as it involves data-science techniques that are not yet common in chemistry and materials science. Here, we provide a tutorial-style overview of strategies and best practices for the construction of artificial neural network (ANN) potentials. We illustrate the most important aspects of (i) data collection, (ii) model selection, (iii) training and validation, and (iv) testing and refinement of ANN potentials on the basis of practical examples. Current research in the areas of active learning and delta learning are also discussed in the context of ANN potentials. This tutorial review aims at equipping computational chemists and materials scientists with the required background knowledge for ANN potential construction and application, with the intention to accelerate the adoption of the method, so that it can facilitate exciting research that would otherwise be challenging with conventional strategies.

arXiv comment19 pages, 11 figures

References in corpus (60)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• Gaussian Error Linear Units (GELUs)
• Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
• On representing chemical environments
• Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanics
• SchNet - a deep learning architecture for molecules and materials
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Molecular Graph Convolutions: Moving Beyond Fingerprints
• Perspective on density functional theory
• Machine Learning Force Fields
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Machine Learning of Accurate Energy-Conserving Molecular Force Fields
• A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
• PhysNet: A Neural Network for Predicting Energies, Forces, Dipole Moments and Partial Charges
• Machine learning for molecular simulation
• Big Data meets Quantum Chemistry Approximations: The $\Delta$-Machine Learning Approach
• Less is more: sampling chemical space with active learning
• DScribe: Library of Descriptors for Machine Learning in Materials Science
• Active learning of linearly parametrized interatomic potentials
• A Fourth-Generation High-Dimensional Neural Network Potential with Accurate Electrostatics Including Non-local Charge Transfer
• Machine Learning Molecular Dynamics for the Simulation of Infrared Spectra
• VAMPnets: Deep learning of molecular kinetics
• How van der Waals interactions determine the unique properties of water
• Nuclear quantum effects enter the mainstream
• SchNetPack: A Deep Learning Toolbox For Atomistic Systems
• Ab initio thermodynamics of liquid and solid water
• Nucleation mechanism for the direct graphite-to-diamond phase transition
• Efficient and Accurate Machine-Learning Interpolation of Atomic Energies in Compositions with Many Species
• FCHL revisited: faster and more accurate quantum machine learning
• Alchemical and structural distribution based representation for improved QML
• Interatomic potentials for ionic systems with density functional accuracy based on charge densities obtained by a neural network
• Hierarchical modeling of molecular energies using a deep neural network
• Automatic Selection of Atomic Fingerprints and Reference Configurations for Machine-Learning Potentials
• WACSF - Weighted Atom-Centered Symmetry Functions as Descriptors in Machine Learning Potentials
• Deep Potential: a general representation of a many-body potential energy surface
• Transferability in Machine Learning for Electronic Structure via the Molecular Orbital Basis
• De novo exploration and self-guided learning of potential-energy surfaces
• Incorporating long-range physics in atomic-scale machine learning
• Extending the Accuracy of the SNAP Interatomic Potential Form
• A Machine Learning Potential for Graphene
• Machine learning enables long time scale molecular photodynamics simulations
• Committee neural network potentials control generalization errors and enable active learning
• Machine-learned multi-system surrogate models for materials prediction
• Machine learning of molecular properties: locality and active learning
• Constructing first-principles phase diagrams of amorphous LixSi using machine-learning-assisted sampling with an evolutionary algorithm
• Coupled cluster theory in materials science
• The Interplay of Structure and Dynamics in the Raman Spectrum of Liquid Water over the Full Frequency and Temperature Range
• Predicting the Activity and Selectivity of Bimetallic Metal Catalysts for Ethanol Reforming using Machine Learning
• A Universal Density Matrix Functional from Molecular Orbital-Based Machine Learning: Transferability across Organic Molecules
• Gaussian Moments as Physically Inspired Molecular Descriptors for Accurate and Scalable Machine Learning Potentials
• First-principles interatomic potentials for ten elemental metals via compressed sensing
• Neural Message Passing with Edge Updates for Predicting Properties of Molecules and Materials
• Efficient Training of ANN Potentials by Including Atomic Forces via Taylor Expansion and Application to Water and a Transition-Metal Oxide
• High-Dimensional Potential Energy Surfaces for Molecular Simulations
• Convolutional neural networks for atomistic systems
• A Novel Approach to Describe Chemical Environments in High Dimensional Neural Network Potentials
• An assessment of the structural resolution of various fingerprints commonly used in machine learning
• Potential energy surface interpolation with neural networks for instanton rate calculations
• Simple and efficient algorithms for training machine learning potentials to force data

Cited by in corpus (23)

• GPUMD: A package for constructing accurate machine-learned potentials and performing highly efficient atomistic simulations
• How to validate machine-learned interatomic potentials
• Molecular dynamics simulations of heat transport using machine-learned potentials: A mini review and tutorial on GPUMD with neuroevolution potentials
• Ultra-fast interpretable machine-learning potentials
• Universal QM/MM Approaches for General Nanoscale Applications
• Developments and Further Applications of Ephemeral Data Derived Potentials
• Advances in modeling complex materials: The rise of neuroevolution potentials
• ænet-PyTorch: a GPU-supported implementation for machine learning atomic potentials training
• AENET-LAMMPS and AENET-TINKER: Interfaces for Accurate and Efficient Molecular Dynamics Simulations with Machine Learning Potentials
• Energy-conserving molecular dynamics is not energy conserving
• Reducing the cost of neural network potential generation for reactive molecular systems
• Self-learning hybrid Monte Carlo method for isothermal-isobaric ensemble: Application to liquid silica
• Analyzing Atomic Interactions in Molecules as Learned by Neural Networks
• A machine learning potential for simulating infrared spectra of nanosilicate clusters
• Four-Dimensional-Spacetime Atomistic Artificial Intelligence Models
• Random sampling versus active learning algorithms for machine learning potentials of quantum liquid water
• Physics-informed active learning for accelerating quantum chemical simulations
• Generating candidates in global optimization algorithms using complementary energy landscapes
• Symmetry-invariant quantum machine learning force fields
• Tell machine learning potentials what they are needed for: Simulation-oriented training exemplified for glycine
• Building Robust Machine Learning Models for Small Chemical Science Data: The Case of Shear Viscosity
• Molecular dynamics-driven global tetra-atomic potential energy surfaces: Application to the AlF dimer
• Machine-Learned Potentials for Solvation Modeling