Two-Parametric Compound Binomial Approximations |
| |
Authors: | V. Cekanavicius B. Roos |
| |
Affiliation: | (1) Vilnius University, Naugarduko 24, LT-03225 Vilnius, Lithuania;(2) Department of Mathematics, SPST, University of Hamburg, Bundesstr. 55, 20146 Hamburg, Germany |
| |
Abstract: | We consider two-parametric compound binomial approximation of the generalized Poisson binomial distribution. We show that the accuracy of approximation essentially depends on the symmetry or shifting of distributions and construct asymptotic expansions. For the proofs, we combine the properties of norms with the results for convolutions of symmetric and shifted distributions. In the lattice case, we use the characteristic function method. In the case of almost binomial approximation, we apply Steins method.__________Published in Lietuvos Matematikos Rinkinys, Vol. 44, No. 4, pp. 443–466, October–December, 2004. |
| |
Keywords: | compound binomial distribution concentration function Kolmogorov norm shifted distributions Stein /content/j5312p2r43400n75/xxlarge8217.gif" alt=" rsquo" align=" BASELINE" BORDER=" 0" >s method symmetric distributions total variation norm |
本文献已被 SpringerLink 等数据库收录! |
|