English

On the Gaussian surface area of spectrahedra

Probability 2022-02-03 v2

Abstract

We show that for sufficiently large n1n\geq 1 and d=Cn3/4d=C n^{3/4} for some universal constant C>0C>0, a random spectrahedron with matrices drawn from Gaussian orthogonal ensemble has Gaussian surface area Θ(n1/8)\Theta(n^{1/8}) with high probability.

Keywords

Cite

@article{arxiv.2112.01463,
  title  = {On the Gaussian surface area of spectrahedra},
  author = {Srinivasan Arunachalam and Oded Regev and Penghui Yao},
  journal= {arXiv preprint arXiv:2112.01463},
  year   = {2022}
}

Comments

8 pages. v2: minor changes to improve presentation