Neural networks as a tool for compact representation of ab initio molecular potential energy surfaces

Tafeit, Erwin; Estelberger, Willibald; Horejsi, Renate; Moeller, Reinhard; Oettl, Karl; Vrecko, Karoline; Reibnegger, Gilbert

doi:10.1016/0263-7855(95)00087-9

Cited by 31 publications

(27 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…114 The training and validation sets have been quite small containing 81 and 64 points, respectively, and the fitting errors have been very large. However, the authors could show that a smaller energy range could be fitted with a much improved accuracy.…”

Section: 1 Overviewmentioning

confidence: 99%

Neural network potential-energy surfaces in chemistry: a tool for large-scale simulations

Behler

2011

Phys. Chem. Chem. Phys.

713

633

View full text Add to dashboard Cite

The accuracy of the results obtained in molecular dynamics or Monte Carlo simulations crucially depends on a reliable description of the atomic interactions. A large variety of efficient potentials has been proposed in the literature, but often the optimum functional form is difficult to find and strongly depends on the particular system. In recent years, artificial neural networks (NN) have become a promising new method to construct potentials for a wide range of systems. They offer a number of advantages: they are very general and applicable to systems as different as small molecules, semiconductors and metals; they are numerically very accurate and fast to evaluate; and they can be constructed using any electronic structure method. Significant progress has been made in recent years and a number of successful applications demonstrate the capabilities of neural network potentials. In this Perspective, the current status of NN potentials is reviewed, and their advantages and limitations are discussed.

show abstract

Section: 1 Overviewmentioning

confidence: 99%

Neural network potential-energy surfaces in chemistry: a tool for large-scale simulations

Behler

2011

Phys. Chem. Chem. Phys.

713

633

View full text Add to dashboard Cite

show abstract

“…The earliest NN-based PESs directly use a set of internal coordinates, e.g. distances and angles, as input for the NN [85][86][87][88][89]. However, such approaches have the disadvantage that swapping symmetry equivalent atoms may also change the numerical values of the internal coordinates.…”

Section: Artificial Neural Networkmentioning

confidence: 99%

High-dimensional potential energy surfaces for molecular simulations: from empiricism to machine learning

Unke

Koner

Patra

et al. 2020

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

An overview of computational methods to describe high-dimensional potential energy surfaces suitable for atomistic simulations is given. Particular emphasis is put on accuracy, computability, transferability and extensibility of the methods discussed. They include empirical force fields, representations based on reproducing kernels, using permutationally invariant polynomials, neural network-learned representations and combinations thereof. Future directions and potential improvements are discussed primarily from a practical, application-oriented perspective.

show abstract

“…This group successfully proceeded to perform diffusion Monte Carlo simulations on the HF–HCl system using their neural network model. In a contrasting example, Reibnegger et al 24 attempted to describe the dependence of the potential energy of a mainly rigid tetrahydrobiopterin molecule upon variation of two specific torsional angles within that molecule. A particularly low number of examples were presented to the neural network during the training phase resulting in poor predictions of the potential energy.…”

Section: Review Of the Use Of Neural Network To Design Potentialsmentioning

confidence: 99%

A polarizable high‐rank quantum topological electrostatic potential developed using neural networks: Molecular dynamics simulations on the hydrogen fluoride dimer

Houlding

Liem

Popelier

2007

Int J of Quantum Chemistry

View full text Add to dashboard Cite

ABSTRACT:The HF molecule is a simple polar system that serves as a prototype for developing new potentials. Here we build on earlier work [Liem and Popelier, J Chem Phys 2003, 119, 4560] in which a high-rank multipolar potential was used to simulate liquid HF. That work was the first example of high-rank multipole moments (up to hexadecapole) being employed in conjunction with multipolar Ewald summation in a molecular dynamics simulation. This potential is now extended with polarization, which is delivered by artificial neural networks. The neural nets predict how atomic multipole moments change as the position of neighboring molecules vary. This novel approach is successfully tested on the HF dimer in vacuum.

show abstract

Neural networks as a tool for compact representation of ab initio molecular potential energy surfaces

Cited by 31 publications

References 7 publications

Neural network potential-energy surfaces in chemistry: a tool for large-scale simulations

Neural network potential-energy surfaces in chemistry: a tool for large-scale simulations

High-dimensional potential energy surfaces for molecular simulations: from empiricism to machine learning

A polarizable high‐rank quantum topological electrostatic potential developed using neural networks: Molecular dynamics simulations on the hydrogen fluoride dimer

Contact Info

Product

Resources

About