Reconstructing Kernel-based Machine Learning Force Fields with Super-linear Convergence
| Autor: | Blücher, Stefan, Müller, Klaus-Robert, Chmiela, Stefan |
|---|---|
| Rok vydání: | 2022 |
| Předmět: | |
| Zdroj: | Journal of Chemical Theory and Computation (2023) |
| Druh dokumentu: | Working Paper |
| DOI: | 10.1021/acs.jctc.2c01304 |
| Popis: | Kernel machines have sustained continuous progress in the field of quantum chemistry. In particular, they have proven to be successful in the low-data regime of force field reconstruction. This is because many equivariances and invariances due to physical symmetries can be incorporated into the kernel function to compensate for much larger datasets. So far, the scalability of kernel machines has however been hindered by its quadratic memory and cubical runtime complexity in the number of training points. While it is known, that iterative Krylov subspace solvers can overcome these burdens, their convergence crucially relies on effective preconditioners, which are elusive in practice. Effective preconditioners need to partially pre-solve the learning problem in a computationally cheap and numerically robust manner. Here, we consider the broad class of Nystr\"om-type methods to construct preconditioners based on successively more sophisticated low-rank approximations of the original kernel matrix, each of which provides a different set of computational trade-offs. All considered methods aim to identify a representative subset of inducing (kernel) columns to approximate the dominant kernel spectrum. Comment: Improving readablity and presenetation of manuscript. 18 pages, 12 figures, preprint |
| Databáze: | arXiv |
| Externí odkaz: |
|