Strong Elliptic Law (Q4396352)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 1165651
Language Label Description Also known as
default for all languages
No label defined
    English
    Strong Elliptic Law
    scientific article; zbMATH DE number 1165651

      Statements

      Strong Elliptic Law (English)
      0 references
      0 references
      14 June 1998
      0 references
      random matrix
      0 references
      eigenvalue
      0 references
      spectral function
      0 references
      canonical equation
      0 references
      Hermitian matrix
      0 references
      \(V\)-transform
      0 references
      strong circular law
      0 references
      strong elliptic law
      0 references
      Let \(\Xi=(\xi_{ij}^{(n)})\), \(i,j=1,\dots,n\), be a random non-selfadjoint random matrix whose pairs of entries \(\{\xi_{ij}^{(n)},\xi_{ji}^{(n)}\}\), \(i>j\), \(i,j=1,\dots,n\), are independent for every \(n\), \(E\xi_{ij}^{(n)}=0\) and let \(\lambda_1,\dots,\lambda_n\) be eigenvalues of \(\Xi\). Denote by \[ \nu_n(B)=n^{-1}\sum_{k=1}^n \chi\{\omega:\lambda_k\in B\}, \quad \nu_n(x,y)=n^{-1}\sum_{k=1}^n \chi\{\text{Re } \lambda_k<x\}\chi\{\text{Im } \lambda_k<y\} \] the normalized spectral functions, where \(\chi(\cdot)\) is the indicator function. Under some additional conditions on the densities of the real and the imaginary parts of \(\sqrt{n}\xi_{ij}^{(n)}\) and moment restrictions on \(\xi_{ij}^{(n)}\) it is possible to prove almost sure convergence of properly centered \(\nu_n(B)\) and \(\nu_n(x,y)\), i.e. the so-called strong circular and strong elliptic laws for \(\nu_n(B)\) and \(\nu_n(x,y)\), respectively. The author gives a historical review on the problem beginning from his papers published in the eighties. In the present paper weaker conditions for the fulfilment of the strong circular and strong elliptic laws are proposed as well as more general methods of proving. The methods are based on the modified \(V\)-transform of spectral functions, substitution of the determinant of the Gramm matrix, invariance principle for random matrices and the so-called reform method introduced by the author.
      0 references

      Identifiers