Title: Doubly smoothed maximum likelihood estimation with application to semiparametric structural measurement error models
Authors: Byungtae Seo - Sungkyunkwan University (Korea, South) [presenting]
Abstract: The maximum likelihood estimator of structural semiparametric measurement error models is known to be inconsistent when there exist additional error-free covariates. We show that the doubly-smoothed maximum likelihood method can resolve this inconsistency. In addition, we propose a partially doubly-smoothed maximum likelihood method that gives a more efficient estimator than fully smoothed estimators. The universal consistency of the proposed method is discussed along with theoretical and numerical examples.