Template-Type: ReDIF-Paper 1.0
Author-Name: Kaymak, U.
Author-Name-Last: Kaymak
Author-Name-First: Uzay
Author-Person: pka115
Author-Name: Ben-David, A.
Author-Name-Last: Ben-David
Author-Name-First: Arie
Author-Name: Potharst, R.
Author-Name-Last: Potharst
Author-Name-First: Rob
Title: AUK: a simple alternative to the AUC
Abstract: The area under Receiver Operating Characteristic (ROC) curve, also known as the AUC-index, is commonly used for ranking the performance of data mining models. The AUC has many merits, such as objectivity and ease of interpretation. However, since it is class indifferent, its usefulness while dealing with highly skewed data sets is questionable, to say the least. In this paper, we propose a simple alternative scalar measure to the AUCindex, the Area Under an Kappa curve (AUK). The proposed AUK-index compensates for the above basic flaw of the AUC by being sensitive to the class distribution. Therefore it is particularly suitable for measuring classifiers’ performance on skewed data sets. After introducing the AUK we explore its mathematical relationship with the AUC and show that there is a nonlinear relation between them.
Creation-Date: 2010-06-01
File-URL: https://repub.eur.nl/pub/19678/ERS-2010-024-LIS.pdf
File-Format: application/pdf
Series: RePEc:ems:eureri
Number: ERS-2010-024-LIS
Classification-JEL: L23, M, M11, R4
Keywords: AUC, AUK, H-measure, Kappa index, ROC curve, area under ROC curve, model ranking, model selection
Handle: RePEc:ems:eureri:19678