Estimation and testing for semiparametric mixtures of partially linear models |
| |
Authors: | Xing Wu Tian Liu |
| |
Affiliation: | School of Statistics and Management, Shanghai University of Finance and Economics, Shanghai, P. R. China |
| |
Abstract: | In this paper, we study the estimation and inference for a class of semiparametric mixtures of partially linear models. We prove that the proposed models are identifiable under mild conditions, and then give a PL–EM algorithm estimation procedure based on profile likelihood. The asymptotic properties for the resulting estimators and the ascent property of the PL–EM algorithm are investigated. Furthermore, we develop a test statistic for testing whether the non parametric component has a linear structure. Monte Carlo simulations and a real data application highlight the interest of the proposed procedures. |
| |
Keywords: | EM algorithm hypothesis testing mixture of regression models partially linear models profile likelihood. |
|
|