Neural Network Models of Potential Energy Surfaces: Prototypical Examples
Authors:
- 21 October 2004
- journal article
- Published by American Chemical Society (ACS) in Journal of Chemical Theory and Computation
- Vol. 1 (1) , 14-23
Abstract
Neural networks can be used generate potential energy hypersurfaces by fitting to a data set of energies at discrete geometries, as might be obtained from ab initio calculations. Prior work has shown that this method can generate accurate fits in complex systems of several dimensions. The present paper explores fundamental properties of neural network potential representations in some simple prototypes of one, two, and three dimensions. Optimal fits to the data are achieved by adjusting the network parameters using an extended Kalman filtering algorithm, which is described in detail. The examples provide insight into the relationships between the form of the function being fit, the amount of data needed for an adequate fit, and the optimal network configuration and number of neurons needed. The quality of the network interpolation is substantially improved if gradients as well as the energy are available for fitting. The fitting algorithm is effective in providing an accurate interpolation of the underlying potential function even when random noise is added to the data used in the fit.Keywords
This publication has 32 references indexed in Scilit:
- A hierarchical family of global analytic Born–Oppenheimer potential energy surfaces for the H+H2 reaction ranging in quality from double-zeta to the complete basis set limitThe Journal of Chemical Physics, 2002
- Parametrized direct dynamics study of rate constants of H with CH4 from 250 to 2400 KThe Journal of Chemical Physics, 2002
- Iterative and direct methods employing distributed approximating functionals for the reconstruction of a potential energy surface from its sampled valuesThe Journal of Chemical Physics, 1999
- Learning to interpolate molecular potential energy surfaces with confidence: A Bayesian approachThe Journal of Chemical Physics, 1999
- On the problem of fitting many-body potentials. I. The minimal maximum error scheme and the paradigm of metal systemsThe Journal of Chemical Physics, 1999
- Molecular dynamics simulation with an ab initio potential energy function and finite element interpolation: The photoisomerization of cis-stilbene in solutionThe Journal of Chemical Physics, 1998
- Potential energy surfaces and reactive dynamics of Zn(3P) with H2The Journal of Chemical Physics, 1996
- Distributed approximating functional approach to fitting and predicting potential surfaces. 1. Atom-atom potentialsChemical Physics Letters, 1996
- A general method for constructing multidimensional molecular potential energy surfaces from ab initio calculationsThe Journal of Chemical Physics, 1996
- An improved H3 potential energy surfaceThe Journal of Chemical Physics, 1991