Impact of Derivative Observations on Gaussian Process Machine Learning Potentials: A Direct Comparison of Three Modeling Approaches

Journal of Chemical Theory and Computation, Год журнала: 2025, Номер unknown

Опубликована: Май 23, 2025

Machine learning (ML) potentials have become a well-established tool for providing inexpensive, yet quantum-mechanically accurate, atomistic simulations. Here, we extend our current modeling procedure, based on Gaussian process regression, to include derivative observations into the ML models. We directly compare three system-energy approaches quantum mechanically derived quantities: (i) atomic energies, (ii) total system energy, and (iii) energy with observations. find that has best performance across board, achieving chemical accuracy fewer training data. In addition, both force errors are around an order of magnitude lower when added models in some cases. follow up discussion multiple advantages proposed method brings, such as improved data set availability ability easily dispersion interactions. Additionally, discuss use cases new approach field FFLUX.

Язык: Английский

---

Section Introduction: The Analysis of Chemical Bonding and the Interpretation of Wave Functions‘Are we there yet?’

Elsevier eBooks, Год журнала: 2023, Номер unknown, С. 1 - 12

Опубликована: Сен. 12, 2023

Язык: Английский

---