摘要
基于多项式样条全局光滑方法,建立函数系数线性自回归模型中系数函数的样条估计.在适当条件下,证明了系数函数多项式样条估计的相合性,并给出了它们的收敛速度.模拟例子验证了理论结果的正确性.
A global smoothing method based on polynomial splines is used to estimate the coefficient functions in functional-coefficient linear autoregressive models. Under some mild conditions, consistency of the polynomial spline estimators is proved. Rates of convergence of these estimators are also given. The main results are verified by a simulation example.
基金
国家自然科学基金(60375003)
航空基础科学基金(03I53059)
关键词
函数系数线性自回归模型
多项式样条估计
相合性
收敛速度
functional-coefficient linear autoregressive model
polynomial spline estimation
consistency
convergence rate.
