Quantile-based Random Kaczmarz for corrupted linear systems of equations

Quantile-based Random Kaczmarz for corrupted linear systems of equations
复制标题

用于损坏线性方程组的基于分位数的随机 Kaczmarz

DOI:
10.1093/imaiai/iaab029
复制
发表时间:
2022
期刊:
Information and Inference: A Journal of the IMA
影响因子:
--
通讯作者:
Steinerberger, Stefan
Steinerberger, Stefan
中科院分区:
--
文献类型:
--
作者:
Steinerberger, Stefan

文献摘要

相似文献

我们考虑线性系统,其中包括规范化的行,,和其中多达项已损坏(可能是任意大的数字)。Haddock,Needell,Rebrova & Swartworth提出了一种基于分位数的Random Kaczmarz方法,并证明了对于某些随机矩阵,它以高似然收敛到真实解。我们证明了一个确定性的版本,构造,任何矩阵,一个numbersuch有收敛的所有扰动。假设一个随机矩阵启发式,这证明了高高斯矩阵的收敛性,最多可达腐败(一个数字,可能会得到改善)。
We consider linear systemswhereconsists of normalized rows,, and where up toentries ofhave been corrupted (possibly by arbitrarily large numbers). Haddock, Needell, Rebrova & Swartworth propose a quantile-based Random Kaczmarz method and show that for certain random matricesit converges with high likelihood to the true solution. We prove a deterministic version by constructing, for any matrix, a numbersuch that there is convergence for all perturbations with. Assuming a random matrix heuristic, this proves convergence for tall Gaussian matrices with up tocorruption (a number that can likely be improved).