Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

LC-GAP: Localized Coulomb Descriptors for the Gaussian Approximation Potential

: Barker, James; Bulin, Johannes; Hamaekers, Jan; Mathias, Sonja


Griebel, Michael; Schüller, Anton; Schweitzer, Marc Alexander:
Scientific Computing and Algorithms in Industrial Simulations : Projects and Products of Fraunhofer SCAI
Cham: Springer International Publishing, 2017
ISBN: 978-3-319-62457-0 (Print)
ISBN: 978-3-319-62458-7 (Online)
Aufsatz in Buch
Fraunhofer SCAI ()

We introduce a novel class of localized atomic environment representations based upon the Coulomb matrix. By combining these functions with the Gaussian approximation potential approach, we present LC-GAP, a new system for generating atomic potentials through machine learning (ML). Tests on the QM7, QM7b and GDB9 biomolecular datasets demonstrate that potentials created with LC-GAP can successfully predict atomization energies for molecules larger than those used for training to chemical accuracy, and can (in the case of QM7b) also be used to predict a range of other atomic properties with accuracy in line with the recent literature. As the best-performing representation has only linear dimensionality in the number of atoms in a local atomic environment, this represents an improvement in both prediction accuracy and computational cost when compared to similar Coulomb matrix-based methods.