Premium
Why Not Try a Robust Regression?
Author(s) -
Hettmansperger THomas P.
Publication year - 1987
Publication title -
australian journal of statistics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.434
H-Index - 41
eISSN - 1467-842X
pISSN - 0004-9581
DOI - 10.1111/j.1467-842x.1987.tb00716.x
Subject(s) - leverage (statistics) , studentized range , studentized residual , regression , robust regression , regression analysis , computer science , simple linear regression , econometrics , statistics , set (abstract data type) , data set , simple (philosophy) , data mining , mathematics , machine learning , programming language , standard error , philosophy , epistemology
Summary We introduce and discuss three important regression diagnostics: leverage, Studentized residuals, and DFFITS. We then develop two approaches to bounded‐influence robust regression based on these diagnostics. The methods are illustrated on a data set using a simple MINITAB program.