Predicting tensorial molecular properties with equivariant machine-learning models

Abstract

Embedding molecular symmetries into machine-learning models is key for efficient learning of chemico-physical scalar properties, but little evidence on how to extend the same strategy to tensorial quantities exists. Here we formulate a scalable equivariant machine-learning model based on local atomic environment descriptors. We apply it to a series of molecules and show that accurate predictions can be achieved for a comprehensive list of dielectric and magnetic tensorial properties of different ranks. These results show that equivariant models are a promising platform to extend the scope of machine learning in materials modelling.

References in corpus (21)

• On representing chemical environments
• Moment Tensor Potentials: a class of systematically improvable interatomic potentials
• A Spectral Analysis Method for Automated Generation of Quantum-Accurate Interatomic Potentials
• Less is more: sampling chemical space with active learning
• Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
• A Fourth-Generation High-Dimensional Neural Network Potential with Accurate Electrostatics Including Non-local Charge Transfer
• SchNetPack: A Deep Learning Toolbox For Atomistic Systems
• Symmetry-Adapted Machine-Learning for Tensorial Properties of Atomistic Systems
• Chemical Shifts in Molecular Solids by Machine Learning
• sGDML: Constructing Accurate and Data Efficient Molecular Force Fields Using Machine Learning
• A complete ab initio view of Orbach and Raman spin-lattice relaxation in a Dysprosium coordination compound
• Accurate Interatomic Force Fields via Machine Learning with Covariant Kernels
• Accurate molecular polarizabilities with coupled-cluster theory and machine learning
• Representations of molecules and materials for interpolation of quantum-mechanical simulations via machine learning
• Deep neural network for the dielectric response of insulators
• Raman Spectrum and Polarizability of Liquid Water from Deep Neural Networks
• Cormorant: Covariant Molecular Neural Networks
• PiNN: A Python Library for Building Atomic Neural Networks of Molecules and Materials
• Equivariant representations for molecular Hamiltonians and N-center atomic-scale properties
• Relevance of Rotationally Equivariant Convolutions for Predicting Molecular Properties
• Thermally Averaged Magnetic Anisotropy Tensors via Machine Learning Based on Gaussian Moments

Cited by in corpus (9)

• Data mining, dashboard and statistical analysis: a powerful framework for the chemical design of molecular nanomagnets
• A spectral-neighbour representation for vector fields: machine-learning potentials including spin
• 3DReact: Geometric deep learning for chemical reactions
• Predicting The One-Particle Density Matrix With Machine Learning
• Lanthanide molecular nanomagnets as probabilistic bits
• Efficient Generation of Stable Linear Machine-Learning Force Fields with Uncertainty-Aware Active Learning
• A Bond-Based Machine Learning Model for Molecular Polarizabilities and A Priori Raman Spectra
• A general formalism for machine-learning models based on multipolar-spherical harmonics
• Representing spherical tensors with scalar-based machine-learning models