Model-free data screening and cleaning



Categories of data screening and cleaning issues are reviewed and discussed. Bivariate model-free approaches, specialized model-free regression techniques, together with order statistic-based (OSB) approaches are illustrated and shown to be of particular value for a variety of data screening and cleaning applications. It is shown how a combination of curve estimation with OSB approaches can help screen a progression to AIDS data set. WIREs Comp Stat 2011 3 168–176 DOI: 10.1002/wics.140

For further resources related to this article, please visit the WIREs website.