Back to results

Virginia Polytechnic Institute and State University

Asymptotic distribution of eigenvalues of random matrices and characterization of the Gaussian distribution by rotational invariance

Abstract

dc:description.abstract

The study falls in the area of random equations; in particular properties of random matrices have been studied. The dissertation makes precise some statistical theories of spectra developed in recent years by a number of physicists. Two basic results have been achieved. The first result is a characterization of the distribution of a symmetric random matrix. Assuming independence of the diagonal and super-diagonal random variables of the symmetric random matrix the following theorem is proved: the distribution of the matrix is invariant under orthogonal similarity transforms if and only if the diagonal random variables are normally distributed with mean μ, and variance 2a², and the off-diagonal elements are normally distributed with mean O and variance a², :for some constants μ, and a² > O. The proof is achieved by solving a functional equation in characteristic functions. This seems to have been first proved in this context by Porter and Rosenzweig (Ann. Acad. Sci. Fennicae. AVI, No. 44, 1960) by a different method and under more restrictive conditions than those given here. The second result deals with the asymptotic distribution of eigenvalues of a synnnetric random matrix as the dimension approaches infinity. Let A<sub>n</sub> be an appropriately normalized n ⨉ n symmetric random matrix and let W<sub>n</sub>(x) denote the empirical distribution function of the eigenvalues of A<sub>n</sub. Under suitable conditions on the random variables of the matrix it is proved that W<sub>n</sub>(x)⟶W(x) as n∞, where W is the absolutely continuous distribution function with a semi-circle density, W(x) = { ⎧ 2/π (1-x²)<sup>1/2</sup>, |x| ≤ 1 ⎨ ⎩ 0 , |x| > 1. The proof is achieved by an intricate combinatorial analysis in conjunction with the method of moments. This result extends a conjecture made by E. P. Wigner ("On the Distribution of the Roots of Certain Symmmetric Matrices," Ann. Math. 67, 1958, 325).

Degree

thesis:*
Name thesis:degree_name
Ph. D.
Level thesis:degree_level
doctoral
Discipline thesis:degree_discipline
Statistics
Department dc:contributor.department
Statistics
Grantor dc:publisher
Virginia Polytechnic Institute and State University
Year dc:date.issued
1970

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Olson, William Howard

Rights

dc:rights
Statement dc:rights
  • In Copyright
Language dc:language.iso
en

Identifiers

dc:identifier.*
Handle dc:identifier.uri
http://hdl.handle.net/10919/110345
OAI identifier oai:identifier
oai:vtechworks.lib.vt.edu:10919/110345

Chain of custody

source
Harvested from
Virginia Tech
Base URL
vtechworks.lib.vt.edu/oai/request
Last updated
2026-07-22
Source record
OAI-PMH GetRecord
related terms
citation

Olson, William Howard. Asymptotic distribution of eigenvalues of random matrices and characterization of the Gaussian distribution by rotational invariance. doctoral thesis, Virginia Polytechnic Institute and State University, 1970. http://hdl.handle.net/10919/110345