M-Estimators in Business Statistics
AbstractRecent years have seen a dynamic development in statistical methods for analysing data contaminated with outliers. One of the more important techniques that can deal with outlying observations is robust regression, which represents four decades of research. Until recently the implementation of robust regression methods, such as M-estimation or MM-estimation, was limited owing to their iterative nature. With advances in computing power and the growing availability of statistical packages, such as R and SAS, Stata, the applicability of robust regression methods has increased considerably.The aim of the study is to evaluate one of these methods, namely M-estimation, using data from a survey of small and medium-sized businesses. The comparison involves nine M-estimators, each based on a different weighting function. The results and conclusions are formulated on the basis of empirical data from the DG-1 business survey.
|Journal series||Statistics in Transition, ISSN 1234-7655, e-ISSN 2450-0291, (B 15 pkt)|
|Publication size in sheets||0.65|
|Keywords in English||robust regression, M-estimation, business statistics, outliers|
|Score|| = 15.0, 20-12-2019, ArticleFromJournal|
= 15.0, 20-12-2019, ArticleFromJournal
|Publication indicators||: 2016 = 0.26|
|Citation count*||2 (2020-06-29)|
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.