Title of article :
PURO: A Package for Unmasking Regression Outliers
Author/Authors :
SINIKSARAN, Enis Istanbul University - Faculty of Political Science, Turkey , SATMAN, Mehmet Hakan Istanbul University - Department of Econometrics, Turkey
Abstract :
Multiple regression outliers should be identified because of their potential effect on parameter estimates and inferences from the regression model. In recent years, researchers have proposed numerous strategies and procedures to identify the outliers. A Mathematica package PURO is introduced which implements seven methods from the latest and most respected outlier detection procedures in the statistics literature.
Keywords :
Regression outlier , Mathematica , Detection procedure , Masking , Swamping , Diagnostic plot.
Journal title :
Gazi University Journal Of Science
Journal title :
Gazi University Journal Of Science