一类半参数回归模型二阶段估计的渐近理论

Asymptotic Theory of Two Stage Estimstor in Semiparamgtric Model

考虑回归模型yi＝xiβ＋g（ti）＋ei，1≤i≤n，g为R1上未知函数，β为p×1维待估参数向量．基于近邻权函数利用文献[1]中给出的方法建立了β和g的估计量βn，gn。

Consider the regression model yi = xiβ + g(ti) + ei for i = 1,2, ..., n. Here g is an unknown function on R1, β is a p × 1 Parameter vectors to be estimated and ei is an unobserved random error. In this paper authors use the method which is suggested in [ 1 ] to establish the useful estimations βn. and βn. based on nearest neighbor weight function. It's shown that in and gn have some nice properties.