讨论了函数系数自回归模型，在误差项服从正则变化尾的情形下，模型的概率性质。
The probabilistic properties of functional coefficient auto-regression models with regularly varying tailed are discussed.
结果表明，动态自回归模型时变参数（时变系数）的变化是有规律的，其增量大体上是一些简单周期函数的叠加。
The results showed that the change of time-varying parameters (coefficient) in dynamic AR model has a regularity. Its increments are piled up by some simple period functions.
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。
A global smoothing method based on polynomial splines is used to estimate the coefficient functions in functional-coefficient linear autoregressive models.
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。
A global smoothing method based on polynomial splines is used to estimate the coefficient functions in functional-coefficient linear autoregressive models.
应用推荐