Partial AUC Maximization from Positive-unlabeled Data
Abstract
The partial area under the receiver operating characteristic curve (pAUC) is an important performance metric for binary classification that summarizes true positive rates within a specific range of false positive rates (FPRs). Classifiers that achieve high pAUC need to be obtained in many real-world applications such as cybersecurity, medical care, and advertising. Although many methods for maximizing the pAUC have been proposed, they typically require both labeled positive and negative data for training. However, in practice, labeled negative data are often difficult to collect due to privacy concerns or the need for high expertise to annotate them. In this paper, we propose a method for maximizing the pAUC from positive and unlabeled (PU) data without negative data. Within an empirical risk minimization framework, we show that the pAUC, including its FPR-dependent thresholds, can be represented using only the positive and marginal densities, and derive an empirical estimator from PU data. A classifier is then trained by maximizing the derived smoothed empirical pAUC estimator. We experimentally demonstrate the effectiveness of the proposed method with ten real-world datasets.
Related
- AUC Maximization from Biased Positive-unlabeled Data with Confidence
- Asymptotically Exact Negative Guidance of Diffusion Models via Positive-Unlabeled Learning
- Fraud Detection via Bayesian Positive-Unlabeled Learning with Gaussian Processes and Multilayer Networks
- When are time-to-event models a waste of time? Bridging mixture cure models and positive-unlabeled learning for binary classification under right-censoring
- Natural coordinates for constrained correlation functions: Partial autocorrelations and the geometry of positive power spectra
BibTeX
@misc{kumagai2026partial,
title = {{Partial AUC Maximization from Positive-unlabeled Data}},
author = {Atsutoshi Kumagai and Tomoharu Iwata and Taishi Nishiyama and Hiroshi Takahashi and Kazuki Adachi and Yasuhiro Fujiwara},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.00284},
eprint = {2610.00284},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.00284},
}
RIS
TY - UNPB
TI - Partial AUC Maximization from Positive-unlabeled Data
AU - Atsutoshi Kumagai
AU - Tomoharu Iwata
AU - Taishi Nishiyama
AU - Hiroshi Takahashi
AU - Kazuki Adachi
AU - Yasuhiro Fujiwara
PY - 2026
DA - 2026/10/01
PB - arXiv
DO - 10.48550/arxiv.2610.00284
UR - https://arxiv.org/abs/2610.00284
AB - The partial area under the receiver operating characteristic curve (pAUC) is an important performance metric for binary classification that summarizes true positive rates within a specific range of false positive rates (FPRs). Classifiers that achieve high pAUC need to be obtained in many real-world applications such as cybersecurity, medical care, and advertising. Although many methods for maximizing the pAUC have been proposed, they typically require both labeled positive and negative data for training. However, in practice, labeled negative data are often difficult to collect due to privacy concerns or the need for high expertise to annotate them. In this paper, we propose a method for maximizing the pAUC from positive and unlabeled (PU) data without negative data. Within an empirical risk minimization framework, we show that the pAUC, including its FPR-dependent thresholds, can be represented using only the positive and marginal densities, and derive an empirical estimator from PU data. A classifier is then trained by maximizing the derived smoothed empirical pAUC estimator. We experimentally demonstrate the effectiveness of the proposed method with ten real-world datasets.
ER -