广义变系数模型的局部非线性似然估计

The locally nonlinear likelihood estimations of the generalized varying-coefficient models

针对广义变系数模型,在局部线性似然估计方法的基础上将关于系数函数的局部线性拟合改进为局部非线性拟合,利用Newton-Raphson迭代解法得到广义变系数模型的局部非线性似然估计,进一步讨论了当连接函数为典则函数时迭代公式的具体结果,并辅以实例.

This paper presents a new improved locally likelihood estimation for fitting the generalized varying-coefficient models by using the local nonlinear fitting means and Newton-Raphson iterative solution.Further discussion and a specific example when the connection functions are canonical functions are established.