Persistent homology-based descriptor for machine-learning potential of amorphous structures
This work addresses the problem of constructing symmetry-invariant descriptors for machine-learning potentials in condensed-matter physics, offering a method that avoids hyperparameter tuning and deep-learning techniques, though it appears incremental as it builds on existing descriptor methods.
The authors tackled the challenge of predicting physical properties of amorphous materials by proposing a novel descriptor based on persistent homology for machine-learning potentials, demonstrating it could predict the average energy per atom of amorphous carbon at various densities with a simple model and showing it has characteristics similar to latent spaces in graph neural networks.
High-accuracy prediction of the physical properties of amorphous materials is challenging in condensed-matter physics. A promising method to achieve this is machine-learning potentials, which is an alternative to computationally demanding ab initio calculations. When applying machine-learning potentials, the construction of descriptors to represent atomic configurations is crucial. These descriptors should be invariant to symmetry operations. Handcrafted representations using a smooth overlap of atomic positions and graph neural networks (GNN) are examples of methods used for constructing symmetry-invariant descriptors. In this study, we propose a novel descriptor based on a persistence diagram (PD), a two-dimensional representation of persistent homology (PH). First, we demonstrated that the normalized two-dimensional histogram obtained from PD could predict the average energy per atom of amorphous carbon (aC) at various densities, even when using a simple model. Second, an analysis of the dimensional reduction results of the descriptor spaces revealed that PH can be used to construct descriptors with characteristics similar to those of a latent space in a GNN. These results indicate that PH is a promising method for constructing descriptors suitable for machine-learning potentials without hyperparameter tuning and deep-learning techniques.