Archives

Monotone empirical Bayes test for a truncation parameter distribution using Linex loss
by TaChen Liang

Vol. 1 No. 3 (2006) P.397~P.411

ABSTRACT

This paper deals with a monotone empirical Bayes test $\delta^*_n$ for a truncation parameter distribution using the linex loss. The asymptotic optimality of $\delta^*_n$ is investigated. Under very mild conditions, it is shown that $\delta^*_n$ is asymptotically optimal with a rate of order $n^{-2/3}$. This rate improves the empirical Bayes test $\delta_n^{X\ S}$ of Xu and Shi (2004) in the sense that faster convergence rate is achieved under conditions relatively weaker than that assumed in Xu and Shi (2004).

KEYWORDS
Asymptotic optimality, Bayes risk, linex loss, rate of convergence, regret

MATHEMATICAL SUBJECT CLASSIFICATION 2010
Primary: 62C12

MILESTONES