A1 Vertaisarvioitu alkuperäisartikkeli tieteellisessä lehdessä
Updates to the DScribe library: New descriptors and derivatives
Tekijät: Laakso Jarno, Himanen Lauri, Homm Henrietta, Morooka Eiaki V., Jäger Marc O. J., Todorović Milika, Rinke Patrick
Kustantaja: AIP Publishing
Julkaisuvuosi: 2023
Journal: Journal of Chemical Physics
Lehden akronyymi: J CHEM PHYS
Artikkelin numero: 234802
Vuosikerta: 158
Numero: 23
Sivujen määrä: 8
ISSN: 0021-9606
eISSN: 1089-7690
DOI: https://doi.org/10.1063/5.0151031
Verkko-osoite: https://doi.org/10.1063/5.0151031
Preprintin osoite: https://arxiv.org/abs/2303.14046
Tiivistelmä
We present an update of the DScribe package, a Python library for atomistic descriptors. The update extends DScribe's descriptor selection with the Valle-Oganov materials fingerprint and provides descriptor derivatives to enable more advanced machine learning tasks, such as force prediction and structure optimization. For all descriptors, numeric derivatives are now available in DScribe. For the many-body tensor representation (MBTR) and the Smooth Overlap of Atomic Positions (SOAP), we have also implemented analytic derivatives. We demonstrate the effectiveness of the descriptor derivatives for machine learning models of Cu clusters and perovskite alloys.
We present an update of the DScribe package, a Python library for atomistic descriptors. The update extends DScribe's descriptor selection with the Valle-Oganov materials fingerprint and provides descriptor derivatives to enable more advanced machine learning tasks, such as force prediction and structure optimization. For all descriptors, numeric derivatives are now available in DScribe. For the many-body tensor representation (MBTR) and the Smooth Overlap of Atomic Positions (SOAP), we have also implemented analytic derivatives. We demonstrate the effectiveness of the descriptor derivatives for machine learning models of Cu clusters and perovskite alloys.
Turun yliopiston Tutkimustietojärjestelmän portaali