Estimating the Common Location
作者:
Somesh Kumar,
Divakar Sharma,
期刊:
Statistics
(Taylor Available online 1995)
卷期:
Volume 26,
issue 3
页码: 231-240
ISSN:0233-1888
年代: 1995
DOI:10.1080/02331889508802492
出版商: Gordon & Breach Science Publishers
关键词: AMS 1980 subject classifications;62F10;Common location;Combining estimators;Cohen and Bhattacharya bounds;Uniform distribution
数据来源: Taylor
摘要:
Suppose we have independent random samples of sizesmandnrespectively from two populations characterized by a common location parameterθand unknown scale parameters. Letψxandψybe odd location estimators ofθbased on individual samples. Cohen (1976) suggested a combined estimatorδa, which under mild conditions on the density, is unbiased and improvesψxfor all a such that 0 <a≤a*(m, n), wherea*(m, n) is a constant andm, ndenote the two sample sizes. Bhattacharya (1981) enlarged this class by finding a larger upperboundA(m, n) for ‘a’. In this paper, we further improveA(m, n) and establish dominance ofδ1, over bothψxandψyin cases where Bhattacharya's bound does not help. We also consider, an unbiased estimator different fromδa, and determine conditions oncto improve bothψxandψy. Our results also take into account the cases not covered by Akai (1982).
点击下载:
PDF (341KB)
返 回