Anomaly detection with semi-supervised classification based on risk estimators

TL;DR

This paper introduces two semi-supervised classification methods for anomaly detection that utilize risk estimators, addressing the limitations of traditional one-class methods which assume only normal data in unlabeled sets.

Contribution

It proposes novel shallow and deep semi-supervised anomaly detection techniques based on unbiased and nonnegative risk estimators, with theoretical bounds and parameter selection strategies.

Findings

Effective detection demonstrated through extensive experiments

Theoretical bounds support the proposed methods

Parameter selection techniques improve empirical performance

Abstract

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.