提出了一种叫做逐元估计法的方法用来估计变系数模型中的未知函数和它们的导数构造了一种快速选择估计量窗宽和快速计算大量估计点的方法,推导了估计量的渐近正态性通过Monte Carlo模拟研究了估计量的有限样本性质.
A componentwise procedure is proposed for estimating the unknown functions and their derivatives in varying-coefficient models. A procedure is developed for fast selecting the band- widths of estimators and fast estimating a great deal of points. The asymptotic normality estimators is derived. Finite sample properties of the procedure in the paper are studied through Monte Carlo simulations.