Simultaneous identification of influential subsets and variable selection using quantabilities
dc.contributor.author | Van Vuuren J.O. | |
dc.date.accessioned | 2011-05-15T16:02:31Z | |
dc.date.available | 2011-05-15T16:02:31Z | |
dc.date.issued | 2005 | |
dc.description.abstract | Ideally, outlier detection and variable selection should be dealt with at the same time. Literature, however, proposes only a few practical procedures for handling this problem. A procedure is proposed where dummy variables, variable selection, and a simple selection rule are used to find an optimal model which includes the relevant predictors and excludes the outliers influential on the fit. | |
dc.description.version | Article | |
dc.identifier.citation | South African Statistical Journal | |
dc.identifier.citation | 39 | |
dc.identifier.citation | 1 | |
dc.identifier.issn | 0038271X | |
dc.identifier.uri | http://hdl.handle.net/10019.1/12507 | |
dc.title | Simultaneous identification of influential subsets and variable selection using quantabilities | |
dc.type | Article |