摘要
证明了Gamma分布环境因子的最大似然估计是有偏估计,且其偏差为正,进而导出了Gamma分布环境因子的近似无偏估计.利用Cornish-Fisher展开导出了Gamma分布环境因子的广义置信区间,另外也给了Gamma分布环境因子的Bootstrap-t置信区间.利用模拟方法研究了所给近似无偏估计和区间估计的精度,模拟结果显示所给近似无偏估计和区间估计的精度是相当好的.
It is verified that the maximum likelihood estimation of the environment factor for the gamma distribution is biased,and that its bias is positive.Then an approximate unbiased estimation of the environment factor is derived.Using the Cornish-Fisher expansion,the generalized conˉdence interval of the environment factor is proposed.The bootstrap-t conˉdence interval is also given.The performance of the proposed procedures is evaluated by using Monte Carlo simulation.The simulation results show that the proposed point estimation outperforms the MLE and that the coverage percentages of the proposed generalized conˉdence interval are quite close to the nominal coverage probabilities,even for small sample size and for small shape parameter.
作者
王炳兴
吴方涛
周君兴
WANG Bing-xing;WU Fang-tao;ZHOU Jun-xing(School of Statistics and Mathematics, Zhejiang Gongshang University, Hangzhou 310018, China;School of Data Sciences, Zhejiang University of Finance and Economics, Hangzhou 310018, China)
出处
《高校应用数学学报(A辑)》
CSCD
北大核心
2017年第4期493-500,共8页
Applied Mathematics A Journal of Chinese Universities(Ser.A)
基金
国家自然科学基金(11371322
11671303)
浙江省一流学科A类(浙江工商大学统计学)