English

Effective Condition Number Bounds for Convex Regularization

Numerical Analysis 2019-09-30 v3 Numerical Analysis

Abstract

We derive bounds relating Renegar's condition number to quantities that govern the statistical performance of convex regularization in settings that include the 1\ell_1-analysis setting. Using results from conic integral geometry, we show that the bounds can be made to depend only on a random projection, or restriction, of the analysis operator to a lower dimensional space, and can still be effective if these operators are ill-conditioned. As an application, we get new bounds for the undersampling phase transition of composite convex regularizers. Key tools in the analysis are Slepian's inequality and the kinematic formula from integral geometry.

Keywords

Cite

@article{arxiv.1707.01775,
  title  = {Effective Condition Number Bounds for Convex Regularization},
  author = {Dennis Amelunxen and Martin Lotz and Jake Walvin},
  journal= {arXiv preprint arXiv:1707.01775},
  year   = {2019}
}

Comments

17 pages, 4 figures . arXiv admin note: text overlap with arXiv:1408.3016

R2 v1 2026-06-22T20:39:38.775Z