非线性半参数模型最小二乘核估计的迭代算法

Iterative estimating formulas of least-squares kernel estimator of non-linear semiparametric models

基于非线性半参数模型最小二乘核估计,给出了核估计参数分量和非参数分量估计的构造式,导出了利用高斯-牛顿迭代法的计算公式.实测数据的模拟表示该算法具有较好的拟合精度.

Basing on the nonlinear semiparametric models, this paper focuses on the kernel estimating theory of nonlinear semiparametric models under the Least-Squares principle, including the calculating formulae of the estimation of parametric and nonparametric components, and gives the iterative nonlinear estimating formulae of Kernel Estimator. The calculation results of GPS data show that the fit is good.