基于γ散度的单元水平模型小域稳健估计

在基于抽样调查数据对总体参数进行估计的方法中,小域估计方法能够借助于辅助信息对小样本乃至无样本区域的参数进行有效的估计,并被广泛应用于抽样估计领域。单元水平模型作为小域估计的基本模型之一,是处理单元级别数据估计的有力工具之一。在单元水平模型的应用条件中,需假定区域随机误差和模型随机误差均服从正态分布。然而,在抽样调查中,满足这一条件的调查数据是很少的,尤其是在观测数据中出现离群值时。不满足正态性假设条件下的小域估计量会产生较大的偏差和均方误,因此有必要研究针对正态性假设和离群观测值不敏感的稳健估计方法。通过引入γ散度和γ似然函数,构建了基于单元水平模型的小域稳健估计方法,得到了模型参数的稳健估计和小域目标变量的稳健估计。与现有的稳健估计方法相比,所提新方法能更好地处理区域随机误差和模型随机误差非正态的情形,对于目标变量存在离群观测的情形,具有更好的稳健性,估计均方误更小。在利用模拟数据进行验证中,比较了不同误差分布情形下几类常用估计方法得到的估计量的均方误差,并进一步探究了随着污染分布的方差和比率变化,所得估计量的均方误差变化情形。最后,通过应用于经典的小域估计数据,进一步验证了所提新方法的可行性。

Effects of Gamma Irradiation for Sterilization on Aqueous Dispersions of Length Sorted Carbon Nanotubes

There is currently great interest in the potential use of carbon nanotubes as delivery vessels for nanotherapeutics and other medical applications. However, no data are available on the effects of sterilization methods on the properties of nanotube dispersions, the form in which most medical applications will be processed. Here we show the effects of gamma irradiation from a 60Co source on the dispersion and optical properties of single-wall carbon nanotubes in aqueous dispersion. Samples of different length-refined populations were sealed in ampoules and exposed to a dose of approximately 28 kGy, a level sufficient to ensure sterility of the dispersions. In contrast to literature results for solid-phase nanotube samples, the effects of gamma irradiation on the dispersion and optical properties of the nanotube samples were found to be minimal. Based on these results, gamma irradiation appears sufficiently non-destructive to be industrially useful for the sterilization of nanotube dispersions.