Near-optimal data-independent point locations for radial basis function interpolation (Q1776159): Difference between revisions

Radial basis functions are a standard tool for the reconstruction of multivariate functions from scattered data by interpolation. In a practical application of radial basis functions, it is important to choose the interpolation points properly. Based on a solid theoretical analysis, the authors present two algorithms, both of greedy type, for the concrete computation of the interpolation points. Specifically it is shown that the sets of interpolation points provided by the algorithms are nearly optimal.