search for




 

Distinction of an outlier(s) using difference based regression models
Journal of the Korean Data & Information Science Society 2018;29:339-50
Published online March 31, 2018
© 2018 Korean Data and Information Science Society.

Chun Gun Park1

1Kyonggi University, Department of Mathematics
Correspondence to: Associate professor, Department of Mathematics, Kyonggi University, Gyeonggi-do 16227, Korea. Email: cgpark@kgu.ac.kr
Received February 17, 2018; Revised March 22, 2018; Accepted March 22, 2018.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
In a multiple linear regression model, outliers have a significant adverse effect on statistical inference. So far, various outlier detection techniques have been studied in order to avoid the effects of abnormal values on regression analysis. In spite of these efforts, if the model selection and the anomaly occur as a complex problem, the regression analysis becomes difficult. This study introduces the difference based regression model and the somewhat free outlier search algorithm for model selection (Park and Kim, 2017). In addition, Park and Kim (2017) proposed the algorithm for outlier detection in regression that is under “assumption that outliers always exist in the model”. To overcome the weakness of the assumption, this study adds the robust method of outlier detection for single variable to the algorithm and propose simulation studies.
Keywords : Boxplot, difference based regression model, outlier, quartile, triangle area method.