UNIVERSALITY OF COVARIANCE MATRICES

• 成果类型: Article
• 署名作者: Pillai, Natesh S.; Yin, Jun
• 署名单位: Harvard University; University of Wisconsin System; University of Wisconsin Madison
• 刊物名称: ANNALS OF APPLIED PROBABILITY
• ISSN/ISSBN: 1050-5164
• DOI: 10.1214/13-AAP939
• 发表日期: 2014
• 页码: 935-1001
• 关键词: local eigenvalue statistics; bulk universality; spectral statistics; smallest eigenvalue; limit

摘要

In this paper we prove the universality of covariance matrices of the form H-NxN=X+ X where X is an M x N rectangular matrix with independent real valued entries x(ij) satisfying Ex(ij)=0 and Ex(ij)(2)=1/M, N, M ->infinity. Furthermore it is assumed that these entries have sub-exponential tails or sufficiently high number of moments. We will study the asymptotics in the regime N/M=d(N) is an element of (0,infinity), lim(N ->infinity) d(N)not equal 0, infinity. Our main result is the edge universality of the sample covariance matrix at both edges of the spectrum. In the case lim(N ->infinity)d(N)=1, we only focus on the largest eigenvalue. Our proof is based on a novel version of the Green function comparison theorem for data matrices with dependent entries. En route to proving edge universality, we establish that the Stieltjes transform of the empirical eigenvalue distribution of H is given by the Marcenko-Pastur law uniformly up to the edges of the spectrum with an error of order (N-eta)(-1) where eta is the imaginary part of the spectral parameter in the Stieltjes transform. Combining these results with existing techniques we also show bulk universality of covariance matrices. All our results hold for both real and complex valued entries.