ROBUST ESTIMATION OF VARIANCE COMPONENTS MODEL①
(Institute of Surverying Engineering and Land Information,Central South University of Technology, Changsha 410083, P. R. China)
Abstract: Classical least squares estimation consists of minimizing the sum of the squared residuals of observation. Many authors have produced more robust versions of this estimation by replacing the square by something else, such as the absolute value. These approaches have been generalized, and their robust estimations and influence functions of variance components have been presented. The results may have wide practical and theoretical value.
Key words: influence function robust estimation variance component