Journal Article

Small-sample precision of ROC-related estimates

Blaise Hanczar, Jianping Hua, Chao Sima, John Weinstein, Michael Bittner and Edward R. Dougherty

in Bioinformatics

Volume 26, issue 6, pages 822-830
Published in print March 2010 | ISSN: 1367-4803
Published online February 2010 | e-ISSN: 1460-2059 | DOI:
Small-sample precision of ROC-related estimates

More Like This

Show all results sharing this subject:

  • Bioinformatics and Computational Biology


Show Summary Details


Motivation: The receiver operator characteristic (ROC) curves are commonly used in biomedical applications to judge the performance of a discriminant across varying decision thresholds. The estimated ROC curve depends on the true positive rate (TPR) and false positive rate (FPR), with the key metric being the area under the curve (AUC). With small samples these rates need to be estimated from the training data, so a natural question arises: How well do the estimates of the AUC, TPR and FPR compare with the true metrics?

Results: Through a simulation study using data models and analysis of real microarray data, we show that (i) for small samples the root mean square differences of the estimated and true metrics are considerable; (ii) even for large samples, there is only weak correlation between the true and estimated metrics; and (iii) generally, there is weak regression of the true metric on the estimated metric. For classification rules, we consider linear discriminant analysis, linear support vector machine (SVM) and radial basis function SVM. For error estimation, we consider resubstitution, three kinds of cross-validation and bootstrap. Using resampling, we show the unreliability of some published ROC results.

Availability: Companion web site at


Journal Article.  5580 words.  Illustrated.

Subjects: Bioinformatics and Computational Biology

Full text: subscription required

How to subscribe Recommend to my Librarian

Users without a subscription are not able to see the full content. Please, subscribe or login to access all content.