Anomaly Detection by Recombining Gated Unsupervised Experts

TL;DR

The paper introduces ARGUE, a novel anomaly detection method that combines expert networks with a gated mixture architecture, effectively utilizing various levels of prior knowledge including unsupervised, semi-supervised, and self-supervised settings.

Contribution

ARGUE is a new data-driven anomaly detection approach that integrates dedicated experts and a gating mechanism, applicable across different supervision levels and leveraging prior knowledge.

Findings

Prior knowledge about normal data is as valuable as known anomalies.

ARGUE outperforms traditional methods in various settings.

The method effectively combines expert networks for anomaly detection.

Abstract

Anomaly detection has been considered under several extents of prior knowledge. Unsupervised methods do not require any labelled data, whereas semi-supervised methods leverage some known anomalies. Inspired by mixture-of-experts models and the analysis of the hidden activations of neural networks, we introduce a novel data-driven anomaly detection method called ARGUE. Our method is not only applicable to unsupervised and semi-supervised environments, but also profits from prior knowledge of self-supervised settings. We designed ARGUE as a combination of dedicated expert networks, which specialise on parts of the input data. For its final decision, ARGUE fuses the distributed knowledge across the expert systems using a gated mixture-of-experts architecture. Our evaluation motivates that prior knowledge about the normal data distribution may be as valuable as known anomalies.