Anomaly detection with semi-supervised classification based on risk estimators
This work addresses a key limitation in anomaly detection for applications where data labeling is incomplete or noisy, offering incremental improvements over existing one-class methods.
The paper tackled the problem of anomaly detection in scenarios where unlabeled training data may contain anomalies, proposing two semi-supervised classification methods based on risk estimators to overcome the impractical assumption of purely normal data. The results showed strong effectiveness in experiments, with established error bounds and techniques for parameter selection.
A significant limitation of one-class classification anomaly detection methods is their reliance on the assumption that unlabeled training data only contains normal instances. To overcome this impractical assumption, we propose two novel classification-based anomaly detection methods. Firstly, we introduce a semi-supervised shallow anomaly detection method based on an unbiased risk estimator. Secondly, we present a semi-supervised deep anomaly detection method utilizing a nonnegative (biased) risk estimator. We establish estimation error bounds and excess risk bounds for both risk minimizers. Additionally, we propose techniques to select appropriate regularization parameters that ensure the nonnegativity of the empirical risk in the shallow model under specific loss functions. Our extensive experiments provide strong evidence of the effectiveness of the risk-based anomaly detection methods.