多参数离散指数族参数渐近最优的经验Bayes估计的收敛速度  

The convergence rates of empirical Bayes estimation for the parameters of multi-parameter discrete exponential family

作　　者：杨亚宁[1] 韦来生[1] 

机构地区：[1]中国科技大学,合肥230026

出　　处：《应用概率统计》1995年第1期92-102,共11页Chinese Journal of Applied Probability and Statistics

基　　金：The project is supported by the National Natural Science Foundation of China

摘　　要：本文构造了多参数离散指数族参数的渐近最优的经验Bayes（EB）估计,若记B_n（δ_n,G）为δn的全面Bayes风险,R_G最小Bayes风险,则在某些条件下c_1n~(-1）<R_n（δ_n,G）-R_G≤c_2n^-λ（μ-2）/μ对0<λ<1,μ>2成立,其中c_1,c_2为正的常数,Asymptotically optimal empirical Bayes (EB) estimators δn for the parameters of multi-parameter discrete exponential families are proposed. Under suitable conditions, we prove that, the overall Bayes risk of EB estimatorsconverge to the minimum Bayes risk with rate C1n-1<Rn(δn G) -RGc2n, where 0<λ<1 and μ>2, for some positive constants c1, c2.

关 键 词：收敛速度 离散指数族 贝叶斯估计 EB估计 

分 类 号：O212.1[理学—概率论与数理统计]