Using principal component analysis for neural network high-dimensional potential energy surface - Sorbonne Université
Journal Articles The Journal of Chemical Physics Year : 2020

Using principal component analysis for neural network high-dimensional potential energy surface

Abstract

Potential energy surfaces (PESs) play a central role in our understanding of chemical reactions. Despite the impressive development of efficient electronic structure methods and codes, such computations still remain a difficult task for the majority of relevant systems. In this context, artificial neural networks (NNs) are promising candidates to construct the PES of a wide range of systems. However, the choice of suitable molecular descriptors remains a bottleneck for these algorithms. In this work, we show that a principal components analysis (PCA) is a powerful tool to prepare an optimal set of descriptors and to build an efficient NN: this protocol leads to a substantial improvement of the NNs in learning and predicting a PES. Furthermore, the PCA provides a means to reduce the size of the input space (i.e. number of descriptors) without losing accuracy. As an example, we applied this novel approach to the computation of the high-dimensional PES describing the keto-enol tautomerism reaction occurring in the acetone molecule.
Fichier principal
Vignette du fichier
Casier et al. - 2020 - Using principal component analysis for neural netw.pdf (2.29 Mo) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-02947864 , version 1 (24-09-2020)

Identifiers

Cite

Bastien Casier, Stéphane Carniato, Tsveta Miteva, Nathalie Capron, Nicolas Sisourat. Using principal component analysis for neural network high-dimensional potential energy surface. The Journal of Chemical Physics, 2020, 152 (23), pp.234103. ⟨10.1063/5.0009264⟩. ⟨hal-02947864⟩
62 View
257 Download

Altmetric

Share

More