English

Randomized Complexity of Vector-Valued Approximation

Numerical Analysis 2023-06-27 v1 Numerical Analysis

Abstract

We study the randomized nn-th minimal errors (and hence the complexity) of vector valued approximation. In a recent paper by the author [Randomized complexity of parametric integration and the role of adaption I. Finite dimensional case (preprint)] a long-standing problem of Information-Based Complexity was solved: Is there a constant c>0c>0 such that for all linear problems P\mathcal{P} the randomized non-adaptive and adaptive nn-th minimal errors can deviate at most by a factor of cc? That is, does the following hold for all linear P\mathcal{P} and nNn\in {\mathbb N} \begin{equation*} e_n^{\rm ran-non} (\mathcal{P})\le ce_n^{\rm ran} (\mathcal{P}) \, {\bf ?} \end{equation*} The analysis of vector-valued mean computation showed that the answer is negative. More precisely, there are instances of this problem where the gap between non-adaptive and adaptive randomized minimal errors can be (up to log factors) of the order n1/8n^{1/8}. This raises the question about the maximal possible deviation. In this paper we show that for certain instances of vector valued approximation the gap is n1/2n^{1/2} (again, up to log factors).

Keywords

Cite

@article{arxiv.2306.13697,
  title  = {Randomized Complexity of Vector-Valued Approximation},
  author = {Stefan Heinrich},
  journal= {arXiv preprint arXiv:2306.13697},
  year   = {2023}
}

Comments

14 pages. arXiv admin note: substantial text overlap with arXiv:2306.13471

R2 v1 2026-06-28T11:13:05.814Z