Kybernetika 28 no. 5, 357-382, 1992

Adaptive maximum-likelihood-like estimation in linear models. I. Consistency

Jan Ámos Víšek


An adaptive estimator of regression model coefficients based on maximization of kernel estimate of likelihood is proposed. Its consistency (in Part 1) and asymptotic normality (in Part 2) is proved. An asymptotic representation of the estimate implies also its asymptotic efficiency.


62F35, 62J05, 62F12, 62G35, 62G07