Nonlinear semiparametric AR(1) model with skew-symmetric innovations |
| |
Authors: | Arezo Hajrajabi Afshin Fallah |
| |
Affiliation: | Faculty of Basic Sciences, Imam Khomeini International University, Qazvin, Iran |
| |
Abstract: | In this paper, we expand a first-order nonlinear autoregressive (AR) model with skew normal innovations. A semiparametric method is proposed to estimate a nonlinear part of model by using the conditional least squares method for parametric estimation and the nonparametric kernel approach for the AR adjustment estimation. Then computational techniques for parameter estimation are carried out by the maximum likelihood (ML) approach using Expectation-Maximization (EM) type optimization and the explicit iterative form for the ML estimators are obtained. Furthermore, in a simulation study and a real application, the accuracy of the proposed methods is verified. |
| |
Keywords: | Autoregressive model EM algorithm Maximum likelihood Semiparametric estimation Skew normal innovations |
|
|