{"id":"arxiv/2608.20812","url":"https://arxiv.org/abs/2608.20812","tags":["cs.LG","math.FA"],"score":0,"title":"Resolution-Consistent Greedy Neural Approximation on Infinite-Dimensional Spaces","domain":"cs.LG","source":"arxiv","authors":["Pablo M. Berná","Antonio Falcó","Diego Mondéjar"],"reviews":0,"subject":"cs.LG","summary":"We develop constructive approximation and learning guarantees for shallow neural models with infinite-dimensional inputs observed through finitely many coordinates. The analysis is based on a parameter-normalized neural dictionary and its associated weighted variation class. Within this class, the approximation error separates into a distribution-dependent coordinate-truncation term and a greedy finite-width term. For empirical regression, a fully-corrective greedy procedure yields population guarantees whose statistical complexity is uniform in the retained input resolution. The same…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-08-21","reactions":{}}