非指数分布数族参数的经验Bayes估计的收敛速度

The Convergence Rates of Empirical Bayes Estimator for Nonexponential Distribution Families

在"平方损失"下,研究了非指数分布族参数θ的经验Bayes估计,首先利用概率密度函数的核估计,构造了位置参数的经验Bayes(EB)估计量,在适当的条件下获得了它的收敛速度.

The empirical Bayes （EB） estimator of parametric θ in nonexponential distri- bution families is investigated under square loss Functions, at first, by using kernel-type density estimation. The empirical Bayes estimation rules are constructed and under suit- able conditions, the convergence rates are obtained.