Incorporating electronic information into Machine Learning potential energy surfaces via approaching the ground-state electronic energy as a function of atom-based electronic populations

Abstract

Machine Learning (ML) approximations to Density Functional Theory (DFT) potential energy surfaces (PESs) are showing great promise for reducing the computational cost of accurate molecular simulations, but at present they are not applicable to varying electronic states, and in particular, they are not well suited for molecular systems in which the local electronic structure is sensitive to the medium to long-range electronic environment. With this issue as the focal point, we present a new Machine Learning approach called bpopNN for obtaining efficient approximations to DFT PESs. The methodology is based on approaching the true DFT energy as a function of electron populations on atoms, which may be realized in practice with constrained DFT (CDFT). The new approach creates approximations to this function with deep neural networks. These approximations thereby incorporate electronic information naturally into a ML approach, and optimizing the model energy with respect to populations allows the electronic terms to self-consistently adapt to the environment, as in DFT. We confirm the effectiveness of this approach with a variety of calculations on Li${}_n$H${}_n$ clusters.

28 pages, 8 figures

References in corpus (26)

• Adam: A Method for Stochastic Optimization
• On representing chemical environments
• Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
• SchNet - a deep learning architecture for molecules and materials
• DeePMD-kit: A deep learning package for many-body potential energy representation and molecular dynamics
• Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals
• 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
• Less is more: sampling chemical space with active learning
• Comparing molecules and solids across structural and alchemical space
• Machine-learning based interatomic potential for amorphous carbon
• Finding Density Functionals with Machine Learning
• Edge-functionalized and substitutional doped graphene nanoribbons: electronic and spin properties
• Efficient and Accurate Machine-Learning Interpolation of Atomic Energies in Compositions with Many Species
• Symmetry-Adapted Machine-Learning for Tensorial Properties of Atomistic Systems
• Interatomic potentials for ionic systems with density functional accuracy based on charge densities obtained by a neural network
• A Transferable Machine-Learning Model of the Electron Density
• Electronic Ground State of Higher Acenes
• Incorporating long-range physics in atomic-scale machine learning
• Growth Mechanism and Origin of High $s{p}^3$ Content in Tetrahedral Amorphous Carbon
• Orbital-free Bond Breaking via Machine Learning
• Metadynamics for Training Neural Network Model Chemistries: a Competitive Assessment
• Learning molecular energies using localized graph kernels
• Prediction of Atomization Energy Using Graph Kernel and Active Learning
• Completing density functional theory by machine-learning hidden messages from molecules
• Graph Nets for Partial Charge Prediction