A note on recent proposals for computing l1 estimates

Gilbert W. Bassett; Roger W. Koenker

doi:10.1016/0167-9473(92)90174-E

A note on recent proposals for computing l₁ estimates

Gilbert W. Bassett, Roger W. Koenker

Economics

Research output: Contribution to journal › Article › peer-review

Abstract

Recent methods for computing the least absolute value (l₁) estimate have been proposed. In contrast to the usual linear programming formulation of the l₁ problem, the new methods attempt to use least squares residuals to identify observations whose l₁ residuals are equal to zero. Examples are presented to show that the proposals do not produce estimates that are identical or even necessarily close to l₁, and hence the algorithms cannot be recommended as a method for computing the l₁ estimate.

Original language	English (US)
Pages (from-to)	207-211
Number of pages	5
Journal	Computational Statistics and Data Analysis
Volume	14
Issue number	2
DOIs	https://doi.org/10.1016/0167-9473(92)90174-E
State	Published - Aug 1992

Keywords

Least absolute value
Least squares
Linear programming

ASJC Scopus subject areas

Statistics and Probability
Computational Mathematics
Computational Theory and Mathematics
Applied Mathematics

Online availability

10.1016/0167-9473(92)90174-E

Library availability

Discover UIUC Full Text

Cite this

@article{0070c60c2c564ffdaf06ba490f4f8913,

title = "A note on recent proposals for computing l1 estimates",

abstract = "Recent methods for computing the least absolute value (l1) estimate have been proposed. In contrast to the usual linear programming formulation of the l1 problem, the new methods attempt to use least squares residuals to identify observations whose l1 residuals are equal to zero. Examples are presented to show that the proposals do not produce estimates that are identical or even necessarily close to l1, and hence the algorithms cannot be recommended as a method for computing the l1 estimate.",

keywords = "Least absolute value, Least squares, Linear programming",

author = "Bassett, {Gilbert W.} and Koenker, {Roger W.}",

year = "1992",

month = aug,

doi = "10.1016/0167-9473(92)90174-E",

language = "English (US)",

volume = "14",

pages = "207--211",

journal = "Computational Statistics and Data Analysis",

issn = "0167-9473",

publisher = "Elsevier B.V.",

number = "2",

}

TY - JOUR

T1 - A note on recent proposals for computing l1 estimates

AU - Bassett, Gilbert W.

AU - Koenker, Roger W.

PY - 1992/8

Y1 - 1992/8

N2 - Recent methods for computing the least absolute value (l1) estimate have been proposed. In contrast to the usual linear programming formulation of the l1 problem, the new methods attempt to use least squares residuals to identify observations whose l1 residuals are equal to zero. Examples are presented to show that the proposals do not produce estimates that are identical or even necessarily close to l1, and hence the algorithms cannot be recommended as a method for computing the l1 estimate.

AB - Recent methods for computing the least absolute value (l1) estimate have been proposed. In contrast to the usual linear programming formulation of the l1 problem, the new methods attempt to use least squares residuals to identify observations whose l1 residuals are equal to zero. Examples are presented to show that the proposals do not produce estimates that are identical or even necessarily close to l1, and hence the algorithms cannot be recommended as a method for computing the l1 estimate.

KW - Least absolute value

KW - Least squares

KW - Linear programming

UR - http://www.scopus.com/inward/record.url?scp=38249007752&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=38249007752&partnerID=8YFLogxK

U2 - 10.1016/0167-9473(92)90174-E

DO - 10.1016/0167-9473(92)90174-E

M3 - Article

AN - SCOPUS:38249007752

SN - 0167-9473

VL - 14

SP - 207

EP - 211

JO - Computational Statistics and Data Analysis

JF - Computational Statistics and Data Analysis

IS - 2

ER -