Wybrane metody oceny jakości klasyfikatorów – przegląd i przykłady zastosowań
Selected methods for assessing the performance of classifiers – an overview and examples of applications

Summary/Abstract: The ROC curve and the area under the ROC curve (AUC) are popular measures for evaluating the performance of classification models for binary outcomes. Recently, sev-eral new measures have been proposed to assess the predictive ability of classifiers. These include, among others, reclassification tables [Cook 2008], net reclassification improvement (NRI) and integrated discrimination improvement (IDI) [Pencina et al. 2008]. This paper briefly describes the methods listed above and presents some examples of their application possibilities.

• Details
• Contents