摘要
利用指数分布的若干个样本分位数建立线性回归模型;由获得的指数分布参数的广义最小二乘估计的渐近正态性,得到数据缺失、分组数据、截尾样本场合指数分布参数的渐近估计.在样本足够大的情况下,该方法简单有效.
A linear regression model is established by using several sample quantiles of exponential distribution.By aid of asymptotic normality of the generalized least squares estimation,which obtained by the corresponding parameters of exponential distribution,the approximated interval estimations of exponential distribution parameters are obtained in the cases of the missing data,grouped data and censored sample.The method is simple and effective for large enough sample.
出处
《西北师范大学学报（自然科学版）》
CAS
北大核心
2012年第2期15-18,共4页
Journal of Northwest Normal University(Natural Science)
基金
甘肃省自然科学基金资助项目(096RJZE106)
甘肃省教育厅科研基金资助项目(0908-07)
天水师范学院科研基金资助项目(TSA0931)
关键词
指数分布
样本分位数
广义最小二乘法
渐近正态性
区间估计
exponential distribution
sample quantiles
generalized least squares estimation
asymptotic normality
interval estimation
