Efficient Training of ANN Potentials by Including Atomic Forces via Taylor Expansion and Application to Water and a Transition-Metal Oxide

Abstract

Artificial neural network (ANN) potentials enable the efficient large-scale atomistic modeling of complex materials with near first-principles accuracy. For molecular dynamics simulations, accurate energies and interatomic forces are a prerequisite, but training ANN potentials simultaneously on energies and forces from electronic structure calculations is computationally demanding. Here, we introduce an efficient alternative method for the training of ANN potentials on energy and force information based on an extrapolation of the total energy via a Taylor expansion. By translating the force information to approximate energies, the quadratic scaling with the number of atoms exhibited by conventional force-training methods can be avoided, which enables the training on reference data sets containing complex atomic structures. We demonstrate for different materials systems, clusters of water molecules, bulk liquid water, and a lithium transition-metal oxide, that the proposed force-training approach provides substantial improvements over schemes that train on energies only. Including force information for training reduces the size of the reference data sets required for ANN potential construction, increases the transferability of the potential, and generally improves the force prediction accuracy. For a set of water clusters, the Taylor-expansion approach achieves around 50% of the force error improvement compared to the explicit training on all force components, at a much smaller computational cost. The alternative force training approach thus simplifies the construction of general ANN potentials for the prediction of accurate energies and interatomic forces for diverse types of materials, as demonstrated here for water and a transition-metal oxide.

arXiv comment19 pages, 13 figures

References in corpus (18)

• Canonical sampling through velocity-rescaling
• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• ANI-1: An extensible neural network potential with DFT accuracy at force field computational cost
• Perspective on density functional theory
• A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
• Data-driven materials science: status, challenges and perspectives
• How van der Waals interactions determine the unique properties of water
• Efficient and Accurate Machine-Learning Interpolation of Atomic Energies in Compositions with Many Species
• Density Functional Calculations On First-Row Transition Metals
• A neural network interatomic potential for the phase change material GeTe
• The electronic-structure origin of cation disorder in transition-metal oxides
• Impact of lattice relaxations on phase transitions in a high-entropy alloy studied by machine-learning potentials
• Machine-learned multi-system surrogate models for materials prediction
• Constructing first-principles phase diagrams of amorphous LixSi using machine-learning-assisted sampling with an evolutionary algorithm
• Automated Fitting of Neural Network Potentials at Coupled Cluster Accuracy: Protonated Water Clusters as Testing Ground
• The Interplay of Structure and Dynamics in the Raman Spectrum of Liquid Water over the Full Frequency and Temperature Range
• Potential energy surface interpolation with neural networks for instanton rate calculations
• Low-Temperature Kinetic Isotope Effects in CH3OH+H -> CH2OH+H2 Shed Light on the Deuteration of Methanol in Space

Cited by in corpus (17)

• Machine-learning interatomic potentials for materials science
• A practical guide to machine learning interatomic potentials -- Status and future
• Strategies for the Construction of Machine-Learning Potentials for Accurate and Efficient Atomic-Scale Simulations
• Ephemeral data derived potentials for random structure search
• Data-driven simulation and characterisation of gold nanoparticle melting
• Transfer learning for chemically accurate interatomic neural network potentials
• Predicting Properties of Periodic Systems from Cluster Data: A Case Study of Liquid Water
• Fast and Sample-Efficient Interatomic Neural Network Potentials for Molecules and Materials Based on Gaussian Moments
• ænet-PyTorch: a GPU-supported implementation for machine learning atomic potentials training
• Data-driven Approach to Parameterize SCAN+U for an Accurate Description of 3d Transition Metal Oxide Thermochemistry
• chemtrain: Learning Deep Potential Models via Automatic Differentiation and Statistical Physics
• Constructing and evaluating machine-learned interatomic potentials for Li-based disordered rocksalts
• Self-learning hybrid Monte Carlo method for isothermal-isobaric ensemble: Application to liquid silica
• Fortnet, a software package for training Behler-Parrinello neural networks
• TayFCS: Towards Light Feature Combination Selection for Deep Recommender Systems
• Self-learning path integral hybrid Monte Carlo with mixed ab initio and machine learning potentials for modeling nuclear quantum effects in water
• Scalable Training of Neural Network Potentials for Complex Interfaces Through Data Augmentation