Deep Positive-Unlabeled Anomaly Detection for Contaminated Unlabeled Data
Quick summary
arXiv:2405.18929v3 Announce Type: replace-cross Abstract: Semi-supervised anomaly detection, which aims to improve the anomaly detection performance by using a small amount of labeled anomaly data in addition to unlabeled data, has attracted attention. Existing semi-supervised approaches assume that most unlabeled data are normal, and train anomaly detectors by minimizing the anomaly scores for the unlabeled data while maximizing those for the labeled anomaly data. However, in practice, the unlabeled data are often contaminated with anomalies. This weakens the effect of maximizing the anomaly
Key takeaways
- arXiv:2405.18929v3 Announce Type: replace-cross Abstract: Semi-supervised anomaly detection, which aims to improve the anomaly detection performance by using a small amount of labeled anomaly data in addition to unlabeled data, has attracted attention.
- Existing semi-supervised approaches assume that most unlabeled data are normal, and train anomaly detectors by minimizing the anomaly scores for the unlabeled data while maximizing those for the labeled anomaly data.
- However, in practice, the unlabeled data are often contaminated with anomalies.
Why it matters
The importance of “Deep Positive-Unlabeled Anomaly Detection for Contaminated Unlabeled Data” will be measured by what changes in practice. User behavior, access conditions, verifiable performance and responsible-use outcomes are the signals worth following.
Member comments