Interval-Based AUC(iAUC): Extending ROC Analysis to Uncertainty-Aware Classification
Yuqi Li 1 Matthew M. Engelhard 1 1 Department of Biostatistics & Bioinformatics, Duke University {yuqi.li, m.engelhard}@duke.edu · 2026
Read Interval-Based AUC(iAUC): Extending ROC Analysis to Uncertainty-Aware Classification by Yuqi Li 1 Matthew M. Engelhard 1 1 Department of Biostatistics & Bioinformatics, Duke University {yuqi.li, m.engelhard}@duke.edu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-04775.txt.
Description
Yuqi Li and Matthew M. Engelhard extend a classic evaluation tool in their 2026 work on risk prediction. The ROC curve and its area summarize how well a score separates positive from negative cases, but they assume each prediction is a single number, not a range.
High-stakes predictions often come as intervals that express uncertainty. The authors build two ROC-style curves from interval endpoints and define an interval-based area measure. Rankings split into three regions: confidently correct, ambiguous, and confidently incorrect. Under valid coverage, the new quantities bound the best achievable standard score. Experiments on clinical data showed models with equal ordinary scores can differ sharply in uncertainty-aware ranking.
Catalog record
| CREATOR | Yuqi Li 1 Matthew M. Engelhard 1 1 Department of Biostatistics & Bioinformatics, Duke University {yuqi.li, m.engelhard}@duke.edu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-04775