Smart energy models for atomistic simulations using a DFT-driven multifidelity approach

TL;DR

This paper presents a multifidelity approach for atomistic energy modeling that leverages correlations between low- and high-fidelity data to improve efficiency and reduce the need for extensive high-fidelity calculations.

Contribution

It introduces a multifidelity method for atomistic simulations that outperforms neural networks in data efficiency and computational cost.

Findings

Accurately predicts formation and migration energies in iron-copper alloys.

Requires less training data than neural network models.

Demonstrates comparable or improved accuracy with reduced computational effort.

Abstract

The reliability of atomistic simulations depends on the quality of the underlying energy models providing the source of physical information, for instance for the calculation of migration barriers in atomistic Kinetic Monte Carlo simulations. Accurate (high-fidelity) methods are often available, but since they are usually computationally expensive, they must be replaced by less accurate (low-fidelity) models that introduce some degrees of approximation. Machine-learning techniques such as artificial neural networks are usually employed to work around this limitation and extract the needed parameters from large databases of high-fidelity data, but the latter are often computationally expensive to produce. This work introduces an alternative method based on the multifidelity approach, where correlations between high-fidelity and low-fidelity outputs are exploited to make an educated guess of the high-fidelity outcome based only on quick low-fidelity estimations, hence without the need of running full expensive high-fidelity calculations. With respect to neural networks, this approach is expected to require less training data because of the lower amount of fitting parameters involved. The method is tested on the prediction of ab initio formation and migration energies of vacancy diffusion in iron-copper alloys, and compared with the neural networks trained on the same database.