Asymptotic expansions relating to discrimination based on 2-step monotone missing samples
Nobumichi Shutoha · 2010
This paper provides two asymptotic expansions derived in Okamoto (1963) and McLachlan (1973) based on 2-step monotone missing samples, i.e., certain extensions of Okamoto (1963) and McLachlan (1973) up to the first order. These asymptotic expansions are useful for approximating the probabilities of misclassification and considering of their interval estimation. For investigating the performances in approximating the probabilities, simulation studies for asymptotic expansion of the type of Okamoto (1963) are also given under the selected parameters.