Sharp bounds on the rate of convergence of the empirical covariance matrix

preprint2012arXivOpen access

Abstract

Let X1,...,XN∈RnX_1,..., X_N\in\R^n be independent centered random vectors with log-concave distribution and with the identity as covariance matrix. We show that with overwhelming probability at least 1−3exp⁡(−cn)˚1 - 3 \exp(-c\sqrt{n}\r) one has $ \sup_{x\in S^{n-1}} \Big|\frac{1/N}\sum_{i=1}^N (|<X_i, x>|^2 - \E|<X_i, x>|^2\r)\Big| \leq C \sqrt{\frac{n/N}},$ where CC is an absolute positive constant. This result is valid in a more general framework when the linear forms (<Xi,x>)i≤N,x∈Sn−1(<X_i,x>)_{i\leq N, x\in S^{n-1}} and the Euclidean norms (∣Xi∣/n)i≤N(|X_i|/\sqrt n)_{i\leq N} exhibit uniformly a sub-exponential decay. As a consequence, if AA denotes the random matrix with columns (Xi)(X_i), then with overwhelming probability, the extremal singular values λminλ_{\rm min} and λmaxλ_{\rm max} of AA⊤AA^\top satisfy the inequalities 1−Cn/N≤λmin/N≤λmax/N≤1+Cn/N 1 - C\sqrt{n/N} \le {λ_{\rm min}/N} \le \frac{λ_{\rm max}/N} \le 1 + C\sqrt{n/N} which is a quantitative version of Bai-Yin theorem \cite{BY} known for random matrices with i.i.d. entries.

Explore this paper’s authors, topics and related work.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

Reviews 0

Write a reviewWrite

No reviews yet.

Discussion 0

Add a commentWrite

No comments yet.