Output details
11 - Computer Science and Informatics
University of Bristol
Output title
A coherent interpretation of AUC as a measure of aggregated classification performance
Type
E - Conference contribution
DOI
-
Name of conference/published proceedings
Proceedings of the 28th International Conference on Machine Learning
Volume number
-
Issue number
-
First page of article
657
ISSN of proceedings
-
Year of publication
2011
Number of additional authors
2
Additional information
<24> In an influential paper published in the Machine Learning Journal in 2009, David Hand criticised the widely used AUC performance metric as being fundamentally incoherent. This paper rebuts Hand's claim, demonstrating how it arises from an unrealistic assumption he made, that decision thresholds are set optimally. The paper shows such problems disappear under more realistic assumptions, thus offering a new interpretation.
Interdisciplinary
-
Cross-referral requested
-
Research group
B - Intelligent Systems Laboratory
Citation count
9
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
