A procedure for $\varepsilon $-comparison of means of two normal distributions
Volume 31 / 2004
Applicationes Mathematicae 31 (2004), 155-160
MSC: 62F15, 62J15.
DOI: 10.4064/am31-2-2
Abstract
For two normal distributions $N(\mu _1,\sigma ^2)$ and $N(\mu _2,\sigma ^2)$ the problem is to decide whether $|\mu _1-\mu _2|\leq \varepsilon $ for a given $\varepsilon $. Two decision rules are given: maximin and bayesian for $\sigma ^2$ known and unknown.
