Estimation of the semiparametric additive time series model with first order autoregressive errors is mainly studied.We suppose that the regression function has a parametric framework,and through the local L2-fitting criterion,parametric vector estimators and semiparametric estimators of the regression function can be given under the use of nonparametric kernel function method.Furthermore,under certain regular conditions,the consistency of the estimators is proved.Finally,simulation research and empirical analysis are presented to evaluate the effectiveness and feasibility of the proposed method.
关键词
半参数可加时间序列模型/一阶自回归误差/局部二次拟合准则/核函数调整
Key words
semiparametric additive time series models/first order autoregressive errors/local L2-fitting criterion/kernel function adjustment
维普
万方数据