À propos de cet article
Publié en ligne: 30 déc. 2014
Pages: 61 - 65
DOI: https://doi.org/10.2478/cee-2014-0012
Mots clés
© 2015
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.
Residuals are differences between observed and predicted variables. This paper describes outlier detection method with using studentized internal and external residuals, which was applied to find extreme values in dataset that comes from the planar intersection method. The detected outlier is analysed by the statistical hypothesis testing, where critical value is defined as a quantil of Studentized distribution.