ROBUST ESTIMATION: LOCATION-SCALE AND REGRESSION PROBLEMS

Abstract

A robust estimating method to search for the bounded function minimizing a risk is proposed. The estimators obtained by the proposed method are the generalizations of several classical robust estimators. In addition, a variety of robust estimators can be obtained, including robust one and multidimensional estimators, robust regression and robust model selection criteria for univariate or multivariate data with independent or correlated errors, and robust Bayes estimators. Further, a simulation study is conducted to evaluate the proposed method.