Quasioptimal alternating projections and their use in low-rank approximation of matrices and tensors

TL;DR

This paper analyzes the convergence of inexact alternating projections with quasioptimal metrics for non-convex sets, applying the theory to develop efficient algorithms for low-rank matrix and tensor approximations.

Contribution

It introduces the concept of quasioptimal projections and proves their local linear convergence for super-regular sets, with applications to low-rank approximation problems.

Findings

Quasioptimal alternating projections converge locally and linearly.

Developed fast algorithms for low-rank matrix and tensor approximation.

Numerical experiments show the methods are efficient and useful for regularization.

Abstract

We study the convergence of specific inexact alternating projections for two non-convex sets in a Euclidean space. The $\sigma$-quasioptimal metric projection ($\sigma \geq 1$) of a point $x$ onto a set $A$ consists of points in $A$ the distance to which is at most $\sigma$ times larger than the minimal distance $\mathrm{dist}(x,A)$. We prove that quasioptimal alternating projections, when one or both projections are quasioptimal, converge locally and linearly for super-regular sets with transversal intersection. The theory is motivated by the successful application of alternating projections to low-rank matrix and tensor approximation. We focus on two problems -- nonnegative low-rank approximation and low-rank approximation in the maximum norm -- and develop fast alternating-projection algorithms for matrices and tensor trains based on cross approximation and acceleration techniques. The numerical experiments confirm that the proposed methods are efficient and suggest that they can be used to regularise various low-rank computational routines.