Composite quantile regression for linear errors-in-variables models

Composite quantile regression can be more efficient and sometimes arbitrarily more efficient than least squares for non-normal random errors, and almost as efficient for normal random errors. Therefore, we extend composite quantile regression method to linear errors-in-variables models, and prove the asymptotic normality of the proposed estimators. Simulation results and a real dataset are also given to illustrate our the proposed methods.