Pure density functional for strong correlation and the thermodynamic limit from machine learning

2016/09/13 by Li Li, Thomas E. Baker, Steven R. White +1 · 3 citations
Materials Science · Mathematics · Physics and Astronomy · #Advanced Chemical Physics Studies #Atomic orbital #Density functional theory #Density matrix #Density matrix renormalization group #Limit (mathematics) #Machine Learning in Materials Science #Mathematical analysis #Mathematics #Molecule #Physics #Quantum #Quantum chemical #Quantum chemistry #Quantum many-body systems #Quantum mechanics #Statistical physics #Thermodynamic limit #cond-mat.str-el #physics.chem-ph #physics.comp-ph

paper · pdf · doi:10.1103/physrevb.94.245129

published as Phys. Rev. B 94, 245129 (2016) · 10 figures, 9 pages

arxiv created 2016/09/13 · openalex created_date 2016/09/23 · openalex publication_date 2016/12/21 · arxiv updated 2016/12/28 · openalex updated_date 2026/08/05

Abstract

We use the density-matrix renormalization group, applied to a one-dimensional model of continuum Hamiltonians, to accurately solve chains of hydrogen atoms of various separations and numbers of atoms. We train and test a machine-learned approximation to F[n], the universal part of the electronic density functional, to within quantum chemical accuracy. We also develop a data-driven, atom-centered basis set for densities which greatly reduces the computational cost and accurately represents the physical information in the machine-learning calculation. Our calculation (a) bypasses the standard Kohn-Sham approach, avoiding the need to find orbitals, (b) includes the strong correlation of highly stretched bonds without any specific difficulty (unlike all standard DFT approximations), and (c) is so accurate that it can be used to find the energy in the thermodynamic limit to quantum chemical accuracy.

Citations

• Infinite size density matrix renormalization group, revisited
• Universal variational functionals of electron densities, first-order density matrices, and natural spin-orbitals and solution of the v -representability problem
• Ab initio quantum chemistry using the density matrix renormalization group
• Density-functional theory in one dimension for contact-interacting fermions
• The Thomas-Fermi theory of atoms, molecules and solids
• Orbital-free bond breaking via machine learning
• Fast and Accurate Modeling of Molecular Atomization Energies with Machine Learning
• The density-matrix renormalization group in the age of matrix product states
• The calculation of atomic fields
• The density-matrix renormalization group
• Density-functional theory and strong interactions: Orbital ordering in Mott-Hubbard insulators
• Reference electronic structure calculations in one dimension
• Self-Consistent Equations Including Exchange and Correlation Effects
• Mastering the game of Go with deep neural networks and tree search

Cited by

• Mitigating error cancellation in density functional approximations via machine learning correction
• Predicting fermionic densities using a Projected Quantum Kernel method
• Machine learning of phase transitions in the percolation andXYmodels