ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 1509.06121
18
12

Spectral analysis of the Moore-Penrose inverse of a large dimensional sample covariance matrix

21 September 2015
Taras Bodnar
Holger Dette
Nestor Parolya
ArXivPDFHTML
Abstract

For a sample of nnn independent identically distributed ppp-dimensional centered random vectors with covariance matrix Σn\mathbf{\Sigma}_nΣn​ let S~n\tilde{\mathbf{S}}_nS~n​ denote the usual sample covariance (centered by the mean) and Sn\mathbf{S}_nSn​ the non-centered sample covariance matrix (i.e. the matrix of second moment estimates), where p>np> np>n. In this paper, we provide the limiting spectral distribution and central limit theorem for linear spectral statistics of the Moore-Penrose inverse of Sn\mathbf{S}_nSn​ and S~n\tilde{\mathbf{S}}_nS~n​. We consider the large dimensional asymptotics when the number of variables p→∞p\rightarrow\inftyp→∞ and the sample size n→∞n\rightarrow\inftyn→∞ such that p/n→c∈(1,+∞)p/n\rightarrow c\in (1, +\infty)p/n→c∈(1,+∞). We present a Marchenko-Pastur law for both types of matrices, which shows that the limiting spectral distributions for both sample covariance matrices are the same. On the other hand, we demonstrate that the asymptotic distribution of linear spectral statistics of the Moore-Penrose inverse of S~n\tilde{\mathbf{S}}_nS~n​ differs in the mean from that of Sn\mathbf{S}_nSn​.

View on arXiv
Comments on this paper