Transferability in Machine Learning for Electronic Structure via the Molecular Orbital Basis

Abstract

We present a machine learning (ML) method for predicting electronic structure correlation energies using Hartree-Fock input.The total correlation energy is expressed in terms of individual and pair contributions from occupied molecular orbitals, and Gaussian process regression is used to predict these contributions from a feature set that is based on molecular orbital properties, such as Fock, Coulomb, and exchange matrix elements. With the aim of maximizing transferability across chemical systems and compactness of the feature set, we avoid the usual specification of ML features in terms of atom- or geometry-specific information, such atom/element-types, bond-types, or local molecular structure. ML predictions of MP2 and CCSD energies are presented for a range of systems, demonstrating that the method maintains accuracy while providing transferability both within and across chemical families; this includes predictions for molecules with atom-types and elements that are not included in the training set. The method holds promise both in its current form and as a proof-of-principle for the use of ML in the design of generalized density-matrix functionals.

arXiv comment8 pages, 5 figures

References in corpus (13)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• 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
• Molecular Graph Convolutions: Moving Beyond Fingerprints
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Intrinsic atomic orbitals: An unbiased bridge between quantum theory and chemical concepts
• Big Data meets Quantum Chemistry Approximations: The $\Delta$-Machine Learning Approach
• By-passing the Kohn-Sham equations with machine learning
• Machine Learning of Molecular Electronic Properties in Chemical Compound Space
• Accurate sampling using Langevin dynamics
• Comparison of permutationally invariant polynomials, neural networks, and Gaussian approximation potentials in representing water interactions through many-body expansions
• Gaussian approximation potential modeling of lithium intercalation in carbon nanostructures
• Recognizing molecular patterns by machine learning: an agnostic structural definition of the hydrogen bond

Cited by in corpus (52)

• Machine learning for molecular simulation
• Physics-inspired structural representations for molecules and materials
• Deep neural network solution of the electronic Schrödinger equation
• FCHL revisited: faster and more accurate quantum machine learning
• Autonomous discovery in the chemical sciences part I: Progress
• A Transferable Machine-Learning Model of the Electron Density
• OrbNet: Deep Learning for Quantum Chemistry Using Symmetry-Adapted Atomic-Orbital Features
• Incorporating long-range physics in atomic-scale machine learning
• Perspective on integrating machine learning into computational chemistry and materials science
• Representations of molecules and materials for interpolation of quantum-mechanical simulations via machine learning
• Ab-initio quantum chemistry with neural-network wavefunctions
• A Universal Density Matrix Functional from Molecular Orbital-Based Machine Learning: Transferability across Organic Molecules
• OrbNet Denali: A machine learning potential for biological and organic chemistry with semi-empirical cost and DFT accuracy
• Informing Geometric Deep Learning with Electronic Interactions to Accelerate Quantum Chemistry
• Strategies for the Construction of Machine-Learning Potentials for Accurate and Efficient Atomic-Scale Simulations
• Ground state energy functional with Hartree-Fock efficiency and chemical accuracy
• Regression-clustering for Improved Accuracy and Training Cost with Molecular-Orbital-Based Machine Learning
• Physically inspired deep learning of molecular excitations and photoemission spectra
• Equivariant representations for molecular Hamiltonians and N-center atomic-scale properties
• Putting Density Functional Theory to the Test in Machine-Learning-Accelerated Materials Discovery
• Machine learning configuration interaction for ab initio potential energy curves
• Improved accuracy and transferability of molecular-orbital-based machine learning: Organics, transition-metal complexes, non-covalent interactions, and transition states
• A deep neural network for molecular wave functions in quasi-atomic minimal basis representation
• Machine Learning Prediction of DNA Charge Transport
• Deep Neural Network Computes Electron Densities and Energies of a Large Set of Organic Molecules Faster than Density Functional Theory (DFT)
• Accurate Molecular-Orbital-Based Machine Learning Energies via Unsupervised Clustering of Chemical Space
• An orbital-based representation for accurate Quantum Machine Learning
• Representations and Strategies for Transferable Machine Learning Models in Chemical Discovery
• Accelerating Coupled Cluster Calculations with Nonlinear Dynamics and Shallow Machine Learning
• Analytical Gradients for Molecular-Orbital-Based Machine Learning
• Learning the energy curvature versus particle number in approximate density functionals
• Mean-Field Density Matrix Decompositions
• Learning on-top: regressing the on-top pair density for real-space visualization of electron correlation
• Exploring the design space of machine-learning models for quantum chemistry with a fully differentiable framework
• Data-driven Refinement of Electronic Energies from Two-Electron Reduced-Density-Matrix Theory
• Molecular-orbital-based Machine Learning for Open-shell and Multi-reference Systems with Kernel Addition Gaussian Process Regression
• Low-data deep quantum chemical learning for accurate MP2 and coupled-cluster correlations
• Efficient prediction of 3D electron densities using machine learning
• SPA${}^{\mathrm{H}}$M(a,b): encoding the density information from guess Hamiltonian in quantum machine learning representations
• Machine learning for a finite size correction in periodic coupled cluster theory calculations
• Compressing physical properties of atomic species for improving predictive chemistry
• Equivariant neural network for Green's functions of molecules and materials
• Machine learning method to determine concentrations of structural defects in irradiated materials
• Molecular Dipole Moment Learning via Rotationally Equivariant Gaussian Process Regression with Derivatives in Molecular-orbital-based Machine Learning
• Molecular Energy Learning Using Alternative Blackbox Matrix-Matrix Multiplication Algorithm for Exact Gaussian Process
• Ab initio machine learning in chemical compound space
• Physical machine learning outperforms "human learning" in Quantum Chemistry
• Predicting electronic screening for fast Koopmans spectral functional calculations
• Hybrid localized graph kernel for machine learning energy-related properties of molecules and solids
• Learning to Make Chemical Predictions: the Interplay of Feature Representation, Data, and Machine Learning Algorithms
• Electronic specific heat capacities and entropies from density matrix quantum Monte Carlo using Gaussian process regression to find gradients of noisy data
• Machine learning of electronic structure and atomistic properties from the external potential