摘要
讨论响应变量缺失情况下自适应变系数EV模型的估计问题,利用核估计方法对参数和系数函数进行两步估计,证明估计值的相合性和渐近正态性,数值模拟对所得的两步估计值进行比较,并且得出第二步估计值要优于第一步估计值.
The estimation problem in the adaptive variable coefficient EV model with response variables missing is discussed. By using kernel estimation method, the parameter and the coefficient function are es- timated by two-step estimation, the consistency and the asymptotic normality of the estimated value is proved. The two-step estimated values are compared with numerical simulation and it is found that the second-step estimate is superior to the first-step one.
出处
《兰州理工大学学报》
CAS
北大核心
2015年第2期155-159,共5页
Journal of Lanzhou University of Technology
基金
国家自然科学基金(11261031)
关键词
自适应变系数EV模型
缺失数据
相合性
渐近正态性
adaptive variable coefficient EV model
data missing
consistency
asymptotic normality