A machine learning potential construction based on radial distribution function sampling

Sampling reference data is crucial in machine learning potential (MLP) construction. Inadequate coverage of local configurations in reference data may lead to unphysical behaviors in MLP-based molecular dynamics (MLP-MD) simulations. To address this problem, this study proposes a new on-the-fly reference data sampling method called radial distribution function (RDF)-based data sampling for MLP construction. This method detects and extracts anomalous structures from the trajectories of MLP-MD simulations by focusing on the shapes of RDFs. The detected structures are added to the reference data to improve the accuracy of the MLP. This method allows us to realize a reasonable MLP construction for liquid water with minimal additional data. We prepare data from an H2O molecular cluster system and verify whether the constructed MLPs are practical for bulk water systems. MLP-MD simulations without RDF-based data sampling show unphysical behaviors, such as atomic collisions. In contrast, after applying this method, we obtain MLP-MD trajectories with features, such as RDF shapes and angle distributions, that are comparable to those of ab initio MD simulations. Our simulation results demonstrate that the RDF-based data sampling approach is useful for constructing MLPs that are robust to extrapolations from molecular cluster systems to bulk systems without any specialized know-how. The accuracy of machine learning potentials (MLPs) depends on the quality of training data. MLPs constructed with insufficient data cause unphysical behavior. However, the indiscriminate expansion of the training data is an unwelcome approach from the standpoint of computational cost. This work proposed new training data sampling method, radial distribution function (RDF)-based data sampling, which improves the accuracy of MLPs. The incorporation of new training data via our sampling method led to a notable improvement in the RDFs and other physical properties. Our results demonstrate that this method is useful for constructing MLPs which are robust to extrapolation from molecular cluster system to periodic system. image