Beyond training data: how elemental features enhance ML-based formation energy predictions
Digital Discovery · 2025 · 4 (10) · 2972-2982
DOI: 10.1039/d5dd00182j ↗Kurz gefasst
Fachartikel in der Zeitschrift Digital Discovery, 2025. Verfasst von Hamed Mahdavi und 2 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 3.0). Gefördert von: Division of Materials Research, Directorate for Computer and Information Science and Engineering.
Abstract (Originalsprache)
Element-feature MLP enriches QM-trained GNN interatomic potentials, enabling rapid formation-energy estimates that extrapolate to compounds containing unseen elements; accuracy degrades negligibly when 10% of elements are with held.
Förderung
- Division of Materials Research
- Directorate for Computer and Information Science and Engineering
Zitieren
APA
Mahdavi, H., Honavar, V., & Morgan, D. (2025). Beyond training data: how elemental features enhance ML-based formation energy predictions. Digital Discovery, 4(10), 2972-2982. https://doi.org/10.1039/d5dd00182j
Vancouver
Mahdavi H, Honavar V, Morgan D. Beyond training data: how elemental features enhance ML-based formation energy predictions. Digital Discovery. 2025;4(10):2972-2982. doi:10.1039/d5dd00182j
BibTeX
@article{Mahdavi2025,
author = {Mahdavi, Hamed and Honavar, Vasant and Morgan, Dane},
title = {Beyond training data: how elemental features enhance ML-based formation energy predictions},
journal = {Digital Discovery},
year = {2025},
volume = {4},
number = {10},
pages = {2972--2982},
doi = {10.1039/d5dd00182j}
}
EUVIMED