NANAJul 28, 2017

A Gauss--Newton iteration for Total Least Squares problems

arXiv:1608.0161913 citations
Originality Incremental advance
AI Analysis

For practitioners needing robust solutions to overdetermined linear systems with errors in both A and b, this offers a computationally efficient and globally convergent alternative.

The paper presents a globally convergent Gauss-Newton iteration for solving Total Least Squares problems, requiring only the solution of a rank-one perturbed ordinary least squares problem per iteration.

The Total Least Squares solution of an overdetermined, approximate linear equation $Ax \approx b$ minimizes a nonlinear function which characterizes the backward error. We show that a globally convergent variant of the Gauss--Newton iteration can be tailored to compute that solution. At each iteration, the proposed method requires the solution of an ordinary least squares problem where the matrix $A$ is perturbed by a rank-one term.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes