Towards a Trustworthy Anomaly Detection for Critical Applications through Approximated Partial AUC Loss
This addresses the need for reliable anomaly detection in domains like industrial, medical, or cybersecurity, where missing anomalies can have severe consequences, representing a domain-specific incremental improvement.
The paper tackled the problem of anomaly detection in critical applications where false negatives are unacceptable, by proposing a method that optimizes a specific range of the AUC ROC curve to prevent false negatives, resulting in a TPR of 92.52% at a 20.43% FPR across 6 datasets, with a 4.3% TPR improvement at a 12.2% FPR cost compared to state-of-the-art methods.
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.