English

Isotropic Local Laws for Sample Covariance and Generalized Wigner Matrices

Probability 2015-07-17 v4 Mathematical Physics math.MP

Abstract

We consider sample covariance matrices of the form XXX^*X, where XX is an M×NM \times N matrix with independent random entries. We prove the isotropic local Marchenko-Pastur law, i.e. we prove that the resolvent (XXz)1(X^* X - z)^{-1} converges to a multiple of the identity in the sense of quadratic forms. More precisely, we establish sharp high-probability bounds on the quantity v,(XXz)1wv,wm(z)\langle v, (X^* X - z)^{-1} w \rangle - \langle v,w\rangle m(z), where mm is the Stieltjes transform of the Marchenko-Pastur law and v,wCNv, w \in \mathbb C^N. We require the logarithms of the dimensions MM and NN to be comparable. Our result holds down to scales ImzN1+ϵIm z \geq N^{-1+\epsilon} and throughout the entire spectrum away from 0. We also prove analogous results for generalized Wigner matrices.

Keywords

Cite

@article{arxiv.1308.5729,
  title  = {Isotropic Local Laws for Sample Covariance and Generalized Wigner Matrices},
  author = {Alex Bloemendal and Laszlo Erdos and Antti Knowles and Horng-Tzer Yau and Jun Yin},
  journal= {arXiv preprint arXiv:1308.5729},
  year   = {2015}
}
R2 v1 2026-06-22T01:15:22.899Z