A ROC-based reject rule for dichotomizers

@article{tortorella_prl_2005, abstract = {Many complex classification tasks involve a discrimination between two classes. Since in such cases a classification error could frequently have serious consequences, the classifiers employed should ensure a very high reliability to avoid erroneous decisions. Unfortunately this is difficult to obtain in real situations where the classifier can meet samples very different from those examined in the training phase. Moreover, the cost for a wrong classification can be so high that it is convenient to reject the sample which gives raise to an unreliable result. However, despite its relevance, a reject option specifically devised for dichotomizers (i.e. two-class classifiers) has not been yet proposed. This paper presents a novel reject rule for dichotomizers, based on the Receiver Operating Characteristic curve. The rule minimizes the expected classification cost, defined on the basis of classification and error costs peculiar for the application at hand. Experiments performed with different classifier architectures on several data sets publicly available confirmed the effectiveness of the proposed reject rule.}, author = {Tortorella, Francesco }, citeulike-article-id = {1298477}, journal = {Pattern Recognition Letters}, keywords = {dissertation}, month = {January}, number = {2}, pages = {167--180}, posted-at = {2007-05-15 23:43:17}, priority = {5}, title = {A ROC-based reject rule for dichotomizers}, url = {http://www.sciencedirect.com/science/article/B6V15-4DJBWDC-2/2/69b649b8790c2043bf0792b470c0dbac}, volume = {26}, year = {2005} }

TY - JOUR ID - tortorella_prl_2005 L3 - citeulike-article-id:1298477 TI - A ROC-based reject rule for dichotomizers JF - Pattern Recognition Letters VL - 26 IS - 2 SP - 167 EP - 180 N2 - Many complex classification tasks involve a discrimination between two classes. Since in such cases a classification error could frequently have serious consequences, the classifiers employed should ensure a very high reliability to avoid erroneous decisions. Unfortunately this is difficult to obtain in real situations where the classifier can meet samples very different from those examined in the training phase. Moreover, the cost for a wrong classification can be so high that it is convenient to reject the sample which gives raise to an unreliable result. However, despite its relevance, a reject option specifically devised for dichotomizers (i.e. two-class classifiers) has not been yet proposed. This paper presents a novel reject rule for dichotomizers, based on the Receiver Operating Characteristic curve. The rule minimizes the expected classification cost, defined on the basis of classification and error costs peculiar for the application at hand. Experiments performed with different classifier architectures on several data sets publicly available confirmed the effectiveness of the proposed reject rule. KW - dissertation AU - Tortorella, Francesco PY - 2005/01/15/ UR - http://www.sciencedirect.com/science/article/B6V15-4DJBWDC-2/2/69b649b8790c2043bf0792b470c0dbac ER -