摘要
利用二次光滑估计方法研究纵向数据下部分线性模型的估计问题,给出了二次光滑估计的渐近性质.进一步计算表明,在渐近方差不变的前提下,二次光滑估计的渐近偏差的阶op(h^4)低于局部线性估计的渐近偏差的阶op(h^2),即二次光滑估计的效果优于局部线性估计的效果.利用CD4细胞数数据对二次光滑估计方法进行验证表明,本文所得结果正确.
We used the double smoothing estimation method to discuss the estimation problem of partial linear model,the asymptotic properties of the double smoothing local linear estimators are given.Further calculation show that the order op(h^4)of the asymptotic bias of the double smoothing estimation is lower than the order op(h^2)of the asymptotic bias of the local linear estimation with the same asymptotic variance,that is to say,the double smoothing estimation is better than the local linear estimation.Using CD4 cell number data to verify the double smoothing local linear estimation method,the results obtained in this paper are correct.
作者
李生彪
LI Shengbiao(College of Education,Lanzhou University of Arts and Science,Lanzhou 730000,China)
出处
《延边大学学报(自然科学版)》
CAS
2019年第3期201-207,共7页
Journal of Yanbian University(Natural Science Edition)
基金
甘肃省教育科学“十三五”规划项目(GHB0372)
关键词
纵向数据
二次光滑
部分线性模型
渐近偏差
longitudinal data
double smoothing
partial linear model
asymptotic bias
作者简介
李生彪(1981-),男,副教授,研究方向为数理统计及数学建模.