English

Coverage processes on spheres and condition numbers for linear programming

Probability 2011-06-17 v3 Optimization and Control

Abstract

This paper has two agendas. Firstly, we exhibit new results for coverage processes. Let p(n,m,α)p(n,m,\alpha) be the probability that nn spherical caps of angular radius α\alpha in SmS^m do not cover the whole sphere SmS^m. We give an exact formula for p(n,m,α)p(n,m,\alpha) in the case α[π/2,π]\alpha\in[\pi/2,\pi] and an upper bound for p(n,m,α)p(n,m,\alpha) in the case α[0,π/2]\alpha\in [0,\pi/2] which tends to p(n,m,π/2)p(n,m,\pi/2) when απ/2\alpha\to\pi/2. In the case α[0,π/2]\alpha\in[0,\pi/2] this yields upper bounds for the expected number of spherical caps of radius α\alpha that are needed to cover SmS^m. Secondly, we study the condition number C(A){\mathscr{C}}(A) of the linear programming feasibility problem xRm+1Ax0,x0\exists x\in\mathbb{R}^{m+1}Ax\le0,x\ne0 where ARn×(m+1)A\in\mathbb{R}^{n\times(m+1)} is randomly chosen according to the standard normal distribution. We exactly determine the distribution of C(A){\mathscr{C}}(A) conditioned to AA being feasible and provide an upper bound on the distribution function in the infeasible case. Using these results, we show that E(lnC(A))2ln(m+1)+3.31\mathbf{E}(\ln{\mathscr{C}}(A))\le2\ln(m+1)+3.31 for all n>mn>m, the sharpest bound for this expectancy as of today. Both agendas are related through a result which translates between coverage and condition.

Keywords

Cite

@article{arxiv.0712.2816,
  title  = {Coverage processes on spheres and condition numbers for linear programming},
  author = {Peter Bürgisser and Felipe Cucker and Martin Lotz},
  journal= {arXiv preprint arXiv:0712.2816},
  year   = {2011}
}

Comments

Published in at http://dx.doi.org/10.1214/09-AOP489 the Annals of Probability (http://www.imstat.org/aop/) by the Institute of Mathematical Statistics (http://www.imstat.org)

R2 v1 2026-06-21T09:55:01.860Z