Automatic Selection of Atomic Fingerprints and Reference Configurations for Machine-Learning Potentials

Abstract

Machine learning of atomic-scale properties is revolutionizing molecular modelling, making it possible to evaluate inter-atomic potentials with first-principles accuracy, at a fraction of the costs. The accuracy, speed and reliability of machine-learning potentials, however, depends strongly on the way atomic configurations are represented, i.e. the choice of descriptors used as input for the machine learning method. The raw Cartesian coordinates are typically transformed in "fingerprints", or "symmetry functions", that are designed to encode, in addition to the structure, important properties of the potential-energy surface like its invariances with respect to rotation, translation and permutation of like atoms. Here we discuss automatic protocols to select a number of fingerprints out of a large pool of candidates, based on the correlations that are intrinsic to the training data. This procedure can greatly simplify the construction of neural network potentials that strike the best balance between accuracy and computational efficiency, and has the potential to accelerate by orders of magnitude the evaluation of Gaussian Approximation Potentials based on the Smooth Overlap of Atomic Positions kernel. We present applications to the construction of neural network potentials for water and for an Al-Mg-Si alloy, and to the prediction of the formation energies of small organic molecules using Gaussian process regression.

References in corpus (24)

• Quantum ESPRESSO: a modular and open-source software project for quantum simulations of materials
• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• On representing chemical environments
• Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Comparing molecules and solids across structural and alchemical space
• Machine Learning Unifies the Modelling of Materials and Molecules
• Machine learning prediction errors better than DFT accuracy
• Machine-learning based interatomic potential for amorphous carbon
• Machine Learning of Molecular Electronic Properties in Chemical Compound Space
• Machine Learning Molecular Dynamics for the Simulation of Infrared Spectra
• How van der Waals interactions determine the unique properties of water
• Machine Learning Energies of 2 M Elpasolite (ABC${}_2$D${}_6$) Crystals
• Symmetry-Adapted Machine-Learning for Tensorial Properties of Atomistic Systems
• WACSF - Weighted Atom-Centered Symmetry Functions as Descriptors in Machine Learning Potentials
• A neural network interatomic potential for the phase change material GeTe
• Understanding molecular representations in machine learning: The role of uniqueness and target similarity
• Accurate Interatomic Force Fields via Machine Learning with Covariant Kernels
• Accelerated materials property predictions and design using motif-based fingerprints
• A fingerprint based metric for measuring similarities of crystalline structures
• Microscopic origins of the anomalous melting behaviour of high-pressure sodium
• High Order Path Integrals Made Easy
• Recognizing Local and Global Structural Motifs at the Atomic Scale
• Ab initio Modelling of the Early Stages of Precipitation in Al-6000 Alloys

Cited by in corpus (94)

• Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems
• Big-Data Science in Porous Materials: Materials Genomics and Machine Learning
• Physics-inspired structural representations for molecules and materials
• Machine learning a general purpose interatomic potential for silicon
• Machine learning for electronically excited states of molecules
• Machine-learning interatomic potentials for materials science
• Physically-informed artificial neural networks for atomistic modeling of materials
• A Transferable Machine-Learning Model of the Electron Density
• De novo exploration and self-guided learning of potential-energy surfaces
• Combining SchNet and SHARC: The SchNarc machine learning approach for excited-state dynamics
• Accurate molecular polarizabilities with coupled-cluster theory and machine learning
• On the Completeness of Atomic Structure Representations
• Coarse Graining Molecular Dynamics with Graph Neural Networks
• Atom-Density Representations for Machine Learning
• Unsupervised machine learning in atomistic simulations, between predictions and understanding
• Fast and Accurate Uncertainty Estimation in Chemical Machine Learning
• Learning the electronic density of states in condensed matter
• Predicting molecular dipole moments by combining atomic partial charges and atomic dipoles
• Guest Editorial: Special Topic on Data-enabled Theoretical Chemistry
• Strategies for the Construction of Machine-Learning Potentials for Accurate and Efficient Atomic-Scale Simulations
• Tutorial: How to Train a Neural Network Potential
• Recursive evaluation and iterative contraction of $N$-body equivariant features
• A complete description of thermodynamic stabilities of molecular crystals
• Deep Learning for UV Absorption Spectra with SchNarc: First Steps Towards Transferability in Chemical Compound Space
• Simple machine-learned interatomic potentials for complex alloys
• A Bayesian approach to NMR crystal structure determination
• The Evolution of Machine Learning Potentials for Molecules, Reactions and Materials
• Accurate large-scale simulations of siliceous zeolites by neural network potentials
• Machine learning and excited-state molecular dynamics
• Graph Neural Networks Accelerated Molecular Dynamics
• Ensemble Learning of Coarse-Grained Molecular Dynamics Force Fields with a Kernel Approach
• Machine learning accelerated discovery of corrosion-resistant high-entropy alloys
• An assessment of the structural resolution of various fingerprints commonly used in machine learning
• Efficient implementation of atom-density representations
• Fast Neural Network Approach for Direct Covariant Forces Prediction in Complex Multi-Element Extended Systems
• A reactive neural network framework for water-loaded acidic zeolites
• On Machine Learning Force Fields for Metallic Nanoparticles
• Atomic energy mapping of neural network potential
• Towards Linearly Scaling and Chemically Accurate Global Machine Learning Force Fields for Large Molecules
• Development of a general-purpose machine-learning interatomic potential for aluminum by the physically-informed neural network method
• A New Kind of Atlas of Zeolite Building Blocks
• Simulating solvation and acidity in complex mixtures with first-principles accuracy: the case of CH${}_3$SO${}_3$H and H${}_2$O${}_2$ in phenol
• Machine learning many-body potentials for colloidal systems
• Self-learning Hybrid Monte Carlo: A First-principles Approach
• PyXtal FF: a Python Library for Automated Force Field Generation
• Sensitivity and Dimensionality of Atomic Environment Representations used for Machine Learning Interatomic Potentials
• Development of a physically-informed neural network interatomic potential for tantalum
• Descriptors-free Collective Variables From Geometric Graph Neural Networks
• Finite-temperature materials modeling from the quantum nuclei to the hot electrons regime
• Predicting the Charge Density Response in Metal Electrodes
• ænet-PyTorch: a GPU-supported implementation for machine learning atomic potentials training
• Metadynamics sampling in atomic environment space for collecting training data for machine learning potentials
• UV-Visible Absorption Spectra of Solvated Molecules by Quantum Chemical Machine Learning
• Through the eyes of a descriptor: Constructing complete, invertible descriptions of atomic environments
• Quantum neural networks force fields generation
• Optimal radial basis for density-based atomic representations
• Modeling the Ga/As binary system across temperaturesand compositions from first principles
• Modeling of many-body interactions between elastic spheres through symmetry functions
• Hydrogen diffusion in the lower mantle revealed by machine learning potentials
• Machine Learning for compositional disorder: A Comparison Between Different Descriptors and Machine Learning Frameworks
• Neural Networks Potential from the Bispectrum Component: A Case Study on Crystalline Silicon
• A machine learning potential for simulating infrared spectra of nanosilicate clusters
• High-pressure phase behaviors of titanium dioxide revealed by a $\Delta$-learning potential
• Assessing the persistence of chalcogen bonds in solution with neural network potentials
• Wavelet Scattering Networks for Atomistic Systems with Extrapolation of Material Properties
• Random sampling versus active learning algorithms for machine learning potentials of quantum liquid water
• Machine-learning of atomic-scale properties based on physical principles
• Exploring the design space of machine-learning models for quantum chemistry with a fully differentiable framework
• Predicting hot-electron free energies from ground-state data
• A Hessian-Based Assessment of Atomic Forces for Training Machine Learning Interatomic Potentials
• A deep learning model for chemical shieldings in molecular organic solids including anisotropy
• Modeling of effective interactions between ligand coated nanoparticles through symmetry functions
• Extending the reach of quantum computing for materials science with machine learning potentials
• Size and Quality of Quantum Mechanical Data Sets for Training Neural Network Force Fields for Liquid Water
• Quantum Extreme Learning of molecular potential energy surfaces and force fields
• A smooth basis for atomistic machine learning
• Accelerating the theoretical study of Li-polysulphide adsorption on single-atom catalysts via machine learning approaches
• Electronic Descriptors for Supervised Spectroscopic Predictions
• A Fully Quantum-Mechanical Treatment for Kaolinite
• Neural-networks model for force prediction in multi-principal-element alloys
• A neural network potential with self-trained atomic fingerprints: a test with the mW water potential
• Active-learning-based efficient prediction of ab-initio atomic energy: a case study on a Fe random grain boundary model with millions of atoms
• EOSnet: Embedded Overlap Structures for Graph Neural Networks in Predicting Material Properties
• Critical assessment of machine-learned repulsive potentials for the Density Functional based Tight-Binding method: a case study for pure silicon
• Maximum volume simplex method for automatic selection and classification of atomic environments and environment descriptor compression
• Modeling Chemical Exfoliation of Non-van der Waals Chromium Sulfides by Machine Learning Interatomic Potentials and Monte Carlo Simulations
• Application-specific machine-learned interatomic potentials: exploring the trade-off between DFT convergence, MLIP expressivity, and computational cost
• The Homunculus Brain and Categorical Logic
• Anharmonic infrared spectra of cationic pyrene and superhydrogenated derivatives
• Gaussian Processes for Finite Size Extrapolation of Many-Body Simulations
• Unified theory of atom-centered representations and message-passing machine-learning schemes
• Integer linear programming for unsupervised training set selection in molecular machine learning
• REANN: A PyTorch-based End-to-End Multi-functional Deep Neural Network Package for Molecular, Reactive and Periodic Systems
• Learning electron densities in the condensed phase