Tight error bounds for rank-1 lattice sampling in spaces of hybrid mixed smoothness
Abstract
We consider the approximate recovery of multivariate periodic functions from a discrete set of function values taken on a rank-$s$ integration lattice. The main result is the fact that any (non-)linear reconstruction algorithm taking function values on a rank-$s$ lattice of size $M$ has a dimension-independent lower bound of $2^{-(\alpha+1)/2} M^{-\alpha/2}$ when considering the optimal worst-case error with respect to function spaces of (hybrid) mixed smoothness $\alpha>0$ on the $d$-torus. We complement this lower bound with upper bounds that coincide up to logarithmic terms. These upper bounds are obtained by a detailed analysis of a rank-1 lattice sampling strategy, where the rank-1 lattices are constructed by a component-by-component (CBC) method. This improves on earlier results obtained in [25] and [27]. The lattice (group) structure allows for an efficient approximation of the underlying function from its sampled values using a single one-dimensional fast Fourier transform. This is one reason why these algorithms keep attracting significant interest. We compare our results to recent (almost) optimal methods based upon samples on sparse grids.
- Publication:
-
arXiv e-prints
- Pub Date:
- October 2015
- DOI:
- 10.48550/arXiv.1510.08336
- arXiv:
- arXiv:1510.08336
- Bibcode:
- 2015arXiv151008336B
- Keywords:
-
- Mathematics - Numerical Analysis;
- 65T40;
- 42A10;
- 65D30;
- 65D32;
- 68Q17;
- 68Q25;
- 42B35;
- 65T50;
- 65Y20