Orbital-free bond breaking via machine learning
Autor: | Matthias Rupp, Klaus-Robert Müller, Leo Blooston, John C. Snyder, Kieron Burke, Katja Hansen |
---|---|
Jazyk: | angličtina |
Rok vydání: | 2013 |
Předmět: |
FOS: Computer and information sciences
Electron density physics.chem-ph FOS: Physical sciences General Physics and Astronomy Machine Learning (stat.ML) 02 engineering and technology Bond breaking Machine learning computer.software_genre Kinetic energy 01 natural sciences Molecular dynamics Engineering Statistics - Machine Learning Artificial Intelligence Physics - Chemical Physics 0103 physical sciences Physics::Atomic and Molecular Clusters Computer Simulation Physical and Theoretical Chemistry Physics::Chemical Physics 010306 general physics Computer Science::Databases Chemical Physics (physics.chem-ph) Physics Condensed Matter - Materials Science Chemical Physics business.industry Materials Science (cond-mat.mtrl-sci) 021001 nanoscience & nanotechnology Diatomic molecule stat.ML cond-mat.mtrl-sci Physical Sciences Chemical Sciences Quantum Theory Artificial intelligence 0210 nano-technology business computer Algorithms |
Zdroj: | Snyder, JC; Rupp, M; Hansen, K; Blooston, L; Müller, KR; & Burke, K. (2013). Orbital-free bond breaking via machine learning. Journal of Chemical Physics, 139(22). doi: 10.1063/1.4834075. UC Irvine: Retrieved from: http://www.escholarship.org/uc/item/5rc7m9t2 Snyder, JC; Rupp, M; Hansen, K; Blooston, L; Müller, K-R; & Burke, K. (2013). Orbital-free bond breaking via machine learning. Journal of Chemical Physics, 139(22). doi: 10.1063/1.4834075. UC Irvine: Retrieved from: http://www.escholarship.org/uc/item/0ch480mj The Journal of chemical physics, vol 139, iss 22 The Journal of Chemical Physics |
Popis: | Using a one-dimensional model, we explore the ability of machine learning to approximate the non-interacting kinetic energy density functional of diatomics. This nonlinear interpolation between Kohn-Sham reference calculations can (i) accurately dissociate a diatomic, (ii) be systematically improved with increased reference data and (iii) generate accurate self-consistent densities via a projection method that avoids directions with no data. With relatively few densities, the error due to the interpolation is smaller than typical errors in standard exchange-correlation functionals. © 2013 AIP Publishing LLC. |
Databáze: | OpenAIRE |
Externí odkaz: |