Second order efficiency of maximum likelihood estimators
Second order efficiency of maximum likelihood estimators
复制标题
最大似然估计器的二阶效率
DOI：
--
复制
发表时间：
1974
期刊：
影响因子：
--
通讯作者：
K. Subramanyam
中科院分区：
文献类型：
--
作者：
J. Ghosh;K. Subramanyam
Fisher 和 Rao 的结果扩展到具有多个参数的指数族；即使专门针对 Fisher 和 Rao 的单参数多项式，该证明也是新的。结果应用于伯克森生物测定问题。该论文还提出了 Bhattacharya 型不等式，这是二阶效率结果的（形式）贝叶斯证明以及高达 O(1/n) 的渐近充分性概念。
The results of Fisher and Rao are extended to exponential families with more than one parameter; the proof is new even when specialised to one parameter multinomials of Fisher and Rao. The results are applied to a bioassay problem of Berkson. The paper also develops Bhattacharya type inequalities, a (formal) Bayesian proof of the results on second order efficiency and the notion of asymptotic sufficiency up to O(1/n).