Answers are grounded in this paper and cite their sources.
Active learning enhances efficiency of molecular dynamics simulations, indicating significant cost reductions in training models.
Abstract The use of machine learning (ML) techniques has become increasingly important in computational chemistry and materials science, in recent years. ML potentials can be used for the construction of potential energy surfaces (PES) to avoid computationally expensive ab initio methods. However, many such applications still require a significant number of first-principles calculations to train the ML model, prior to use. Active learning methods can address this issue by performing these calculations and trainings "on-the-fly", based on the ML model’s uncertainty estimation. Nevertheless, current active learning approaches suffer from problems in complex simulations were frequent retraining is required, since repeated training of a large ML model increases training times substantially. This work presents a solution to this limitation by introducing the FALCON (Fast Active Learning for Computational ab initio mOlecular dyNamics) calculator. Instead of relying on a single large ML model, FALCON clusters the training data into subsets of similar structures and distributes them across multiple smaller ML models. This approach significantly increases the efficiency of the OTF training, drastically reducing the computational cost of training-intensive simulations. The use of FALCON is demonstrated on various molecular dynamics (MD) simulations of bulk metals, metal clusters and water diffusion in a carbon nanotube. However, the FALCON calculator is highly flexible and could be easily adapted for various applications and different ML models.
No takes yet. Share an insight, caveat, or question.
Dononelli et al. (2025) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: