SUBSET SELECTION AND THE CONE OF FACTOR-WIDTH-k MATRICES

Research output: Contribution to journalArticlepeer-review

Abstract

We study the cone of factor-width-k matrices, where the factor width of a positive semidefinite matrix is defined as the smallest number k allowing it to be expressed as a sum of positive semidefinite matrices that are nonzero only on a single k\timesk principal submatrix. Two hierarchies of approximations are proposed for this cone. Some theoretical bounds to assess the quality of the new approximations are derived. We also use these approximations to build convex conic relaxations for the subset selection problem where one has to minimize \|b- Ax\|22 under the constraint that x has at most k nonzero components. Several numerical experiments are performed showing that some of these relaxations provide a good compromise between tightness and computational complexity and rank well compared to perspective-type relaxations.

Original languageEnglish
Pages (from-to)817-843
Number of pages27
JournalSIAM Journal on Optimization
Volume34
Issue number1
DOIs
Publication statusPublished - 1 Mar 2024

Keywords

  • cardinality-constrained quadratic optimization
  • conic relaxations
  • factor-width-k matrices
  • subset selection

Fingerprint

Dive into the research topics of 'SUBSET SELECTION AND THE CONE OF FACTOR-WIDTH-k MATRICES'. Together they form a unique fingerprint.

Cite this