针对关键应用设计可信赖的异常检测方法，避免漏检。
Towards a Trustworthy Anomaly Detection for Critical Applications through Approximated Partial AUC Loss
- 用近似部分AUC损失动态优化检测性能
- 在20.43%误报率下实现92.52%真正率
- 适合医疗、工业等对漏报零容忍场景
异常检测是工业、医疗和网络安全等关键领域的关键步骤。这些领域对各类分类错误的处理要求不同：虽可接受误报，但不可容忍漏报，因漏报意味着质量问题、疾病或网络威胁未被发现。为此，我们提出一种动态应用可信近似部分AUC ROC损失（tapAUC）的方法。通过训练二分类器，优化使真正率（TPR）不达100%同时最小化假正率（FPR）的AUC ROC曲线特定区间，最终保留最优阈值用于测试阶段。实验结果表明，在6个数据集上的平均表现达到92.52%的TPR与20.43%的FPR，相比其他先进方法，在假正率增加12.2%的前提下，真正率提升4.3%。代码已开源：https://github.com/ArnaudBougaham/tapAUC。
原文摘要 · Abstract (English)
Anomaly Detection is a crucial step for critical applications such in the industrial, medical or cybersecurity domains. These sectors share the same requirement of handling differently the different types of classification errors. Indeed, even if false positives are acceptable, false negatives are not, because it would reflect a missed detection of a quality issue, a disease or a cyber threat. To fulfill this requirement, we propose a method that dynamically applies a trustworthy approximated partial AUC ROC loss (tapAUC). A binary classifier is trained to optimize the specific range of the AUC ROC curve that prevents the True Positive Rate (TPR) to reach 100% while minimizing the False Positive Rate (FPR). The optimal threshold that does not trigger any false negative is then kept and used at the test step. The results show a TPR of 92.52% at a 20.43% FPR for an average across 6 datasets, representing a TPR improvement of 4.3% for a FPR cost of 12.2% against other state-of-the-art methods. The code is available at https://github.com/ArnaudBougaham/tapAUC.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。