聚集数据线性模型广义聚集双参数改进估计的相对效率

Relative Efficiencies of Improved Estimators of the Generalized Aggregated Double Parameters in the Linear Model with Aggregated Data

针对病态的聚集数据模型,基于Liu型和Stein压缩参数估计方法,提出了聚集数据线性模型广义聚集双参数改进估计的概念,讨论了广义聚集双参数改进估计的优良性,证明了广义聚集双参数改进估计代替最小二乘法估计和Peter&Karsten估计的效率,并得到了两种相对效率的上界.

Based on Liu type and Stein compression estimation method,the concept of an improved estimators of the generalized aggregated double parameters in the linear model for ill-conditioned aggregated data was proposed,and its superiority was discussed.Furthermore,the efficiencies of replacement of the least square estimator and Peter&Karsten estimator with improved estimators of the generalized aggregated double parameters were proved,and the upper bound for two relative efficiencies were obtained.