This article is cited in 1 paper

Asymptotically most accurate double-side confidence intervals in normal-normal model

R. F. Salimov, S. V. Simushkin

Kazan State University, The Faculty of Computer Science and Cybernetics

Abstract: This article presents a form of the most precise class of confidence intervals for the normal-normal model in $d$-posterior approach. It is possible to state its asymptotical behavior with the increase of the sample size. A way of the most asymptotically accurate confidence family construction is also suggested in this article.

Keywords: normal-normal model, $d$-posteriory probability of errors, double-side confidence intervals.

UDC: 517.5+517.9

Received: 09.09.2009

Bibliographic databases:

•