The ROBUSTREG Procedure

Details: ROBUSTREG Procedure

This section describes the statistical and computational aspects of the ROBUSTREG procedure. The following notation is used throughout the section.

Let bold upper X equals left-parenthesis x Subscript i j Baseline right-parenthesis denote an n times p matrix, let bold y equals left-parenthesis y 1 comma ellipsis comma y Subscript n Baseline right-parenthesis prime denote a given n-vector of responses, and let bold-italic theta equals left-parenthesis theta 1 comma ellipsis comma theta Subscript p Baseline right-parenthesis prime denote an unknown p-vector of parameters or coefficients whose components are to be estimated. The matrix bold upper X is called the design matrix. Consider the usual linear model,

bold y equals bold upper X bold-italic theta plus bold e

where bold e equals left-parenthesis e 1 comma ellipsis comma e Subscript n Baseline right-parenthesis prime is an n-vector of unknown errors. It is assumed that (for a given bold upper X) the components e Subscript i of bold e are independent and identically distributed according to a distribution upper L left-parenthesis dot slash sigma right-parenthesis, where sigma is a scale parameter (usually unknown). Often upper L left-parenthesis dot right-parenthesis almost-equals normal upper Phi left-parenthesis dot right-parenthesis, the standard normal distribution function. The vector of residuals for a given value of ModifyingAbove bold-italic theta With caret is denoted by bold r equals left-parenthesis r 1 comma ellipsis comma r Subscript n Baseline right-parenthesis prime, and the ith row of the matrix bold upper X is denoted by bold x prime Subscript i.

Last updated: December 09, 2022