Deep Positive-Unlabeled Anomaly Detection for Contaminated Unlabeled Data

Hiroshi Takahashi, Tomoharu Iwata, Atsutoshi Kumagai, Yuuki Yamanaka

Keywords: Semi-Supervised Anomaly Detection, Positive-Unlabeled Learning, Autoencoder, DeepSVDD

TL;DR: We propose a semi-supervised anomaly detection framework that can effectively handle contaminated unlabeled data.

Abstract:

Semi-supervised anomaly detection has attracted attention,
which aims to improve the anomaly detection performance by using a small amount of labeled anomaly data in addition to unlabeled data.
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 scores for anomalies,
and prevents us from improving the detection performance.
To solve this,
we propose the deep positive-unlabeled anomaly detection framework,
which integrates positive-unlabeled learning with deep anomaly detection models such as autoencoders and deep support vector data descriptions.
Our approach enables the approximation of anomaly scores for normal data using the unlabeled data and the labeled anomaly data.
Therefore,
without labeled normal data,
our approach can train anomaly detectors by minimizing the anomaly scores for normal data while maximizing those for the labeled anomaly data.
Our approach achieves better detection performance than existing approaches on various datasets.

Primary Area: other topics in machine learning (i.e., none of the above)

Submission Number: 14005