Optimality of a utility-based performance metric for ROC analysis

Darrin C. Edwards; Charles E. Metz

doi:10.1117/12.773052

6 March 2008 Optimality of a utility-based performance metric for ROC analysis

Darrin C. Edwards, Charles E. Metz

Proceedings Volume 6917, Medical Imaging 2008: Image Perception, Observer Performance, and Technology Assessment; 69170F (2008) https://doi.org/10.1117/12.773052
Event: Medical Imaging, 2008, San Diego, California, United States

Abstract

We previously introduced a utility-based ROC performance metric, the "surface-averaged expected cost" (SAEC), to address difficulties which arise in generalizing the well-known area under the ROC curve (AUC) to classification tasks with more than two classes. In a two-class classification task, the SAEC can be shown explicitly to be twice the area above the conventional ROC curve (1-AUC) divided by the arclength along the ROC curve. In the present work, we show that for a variety of two-class tasks under the binormal model, the SAEC obtained for the proper decision variable (the likelihood ratio of the latent decision variable) is less than that obtained for the conventional decision variable (i.e., using the latent decision variable directly). We also justify this result using a readily derived property of the arclength along the ROC curve under a given data model. Numerical studies as well as theoretical analysis suggest that the behavior of the SAEC is consistent with that of the AUC performance metric, in the sense that the optimal value of this quantity is achieved by the ideal observer.

Citation Download Citation

Darrin C. Edwards and Charles E. Metz "Optimality of a utility-based performance metric for ROC analysis", Proc. SPIE 6917, Medical Imaging 2008: Image Perception, Observer Performance, and Technology Assessment, 69170F (6 March 2008); https://doi.org/10.1117/12.773052

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available