English

The KLR-theorem revisited

Statistics Theory 2019-02-20 v1 Statistics Theory

Abstract

For independent random variables X1,,Xn;Y1,,YnX_1,\ldots, X_n;Y_1,\ldots, Y_n with all XiX_i identically distributed and same for YjY_j, we study the relation E{aXˉ+bYˉX1Xˉ+Y1Yˉ,,XnXˉ+YnYˉ}=constE\{a\bar X + b\bar Y|X_1 -\bar X +Y_1 -\bar Y,\ldots,X_n -\bar X +Y_n -\bar Y\}={\rm const} with a,ba, b some constants. It is proved that for n3n\geq 3 and ab>0ab>0 the relation holds iff XiX_i and YjY_j are Gaussian.\\ A new characterization arises in case of a=1,b=1a=1, b= -1. In this case either XiX_i or YjY_j or both have a Gaussian component. It is the first (at least known to the author) case when presence of a Gaussian component is a characteristic property.

Cite

@article{arxiv.1902.06800,
  title  = {The KLR-theorem revisited},
  author = {Abram M. Kagan},
  journal= {arXiv preprint arXiv:1902.06800},
  year   = {2019}
}
R2 v1 2026-06-23T07:44:14.036Z