A New Profile Least-Squares Approach for Partially Linear Nonparametric Models with Measurement Errors

Abstract

Partially linear nonparametric models, as a generalization of partially linear models and nonparametric models, have good adaptability and robustness. In this paper, based on the selection of this model, and a new method of contour least squares estimation, the focus is on the estimation of partially linear nonparametric models with measurement errors of covariance. Because when the measurement error is ignored, the parametric and non-parametric components are usually biased estimates. On this basis, we propose a modified contour least squares estimation of measurement error, and also obtain the error variance estimation results. Finally, through numerical simulation, the feasibility of the model proposed in this paper when the covariant has additive measurement error is verified.