Restricted L1 estimators and their covariances /

Bibliographic Details
Main Authors:	Book, D. L. (David Lincoln), 1939- (Author), Booker, J. (Author), Hartley, H. O. (Author), Sielken, Robert L. (Author)
Corporate Authors:	Texas A & M Research Foundation, United States. Office of Naval Research (sponsoring body.)
Format:	Book
Language:	English
Published:	College Station, Texas : Institute of Statistics, Texas A & M University, 1980.
Series:	Technical report (Texas A & M University. Themis Optimization Research Program) ; no. 65.
Subjects:	Analysis of covariance > Computer programs. Parameter estimation > Computer programs. Linear programming. Regression analysis. Computer algorithms. Least squares. Monte Carlo method. Mathematical optimization. Linear constraints Minimizing the sum of absolute residuals Variances and covariances of L1 estimators

Description
Abstract:	The parameters in a linear regression model can be estimated by minimizing the sum of the absolute residuals (L1 estimation) instead of the more classical approach of minimizing the sum of squared residuals (least squares estimation). In addition to other nice properties L1 estimators are less sensitive to outliers than least squares estimators. This paper describes a linear programming algorithm and computer program for obtaining L1 estimators and estimates of their covariances when the regression parameters are restricted to satisfy specified linear constraints. These estimated covariances are the new feature in this work and are an extremely important ingredient in hypothesis tests and confidence interval construction. Technical Report 64 describes a similar procedure for obtaining unbiased L1 estimators when there are no constraints on the parameters.
Item Description:	"June, 1980." "Research conducted through the Texas A & M Research Foundation."
Physical Description:	64 pages, 2 unnumbered pages, 4 pages ; 28 cm
Bibliography:	Includes bibliographical references (page 15).