Double-Adversarial Activation Anomaly Detection: Adversarial Autoencoders are Anomaly Generators
This addresses the problem of detecting anomalies in imbalanced datasets for applications where manual labeling is costly, offering a novel generative approach that improves detection accuracy.
The paper tackles the challenge of unsupervised anomaly detection by introducing DA3D, a method that uses adversarial autoencoders to generate artificial anomalies from normal data, transforming the task into a supervised one and achieving state-of-the-art performance without requiring domain knowledge.
Anomaly detection is a challenging task for machine learning algorithms due to the inherent class imbalance. It is costly and time-demanding to manually analyse the observed data, thus usually only few known anomalies if any are available. Inspired by generative models and the analysis of the hidden activations of neural networks, we introduce a novel unsupervised anomaly detection method called DA3D. Here, we use adversarial autoencoders to generate anomalous counterexamples based on the normal data only. These artificial anomalies used during training allow the detection of real, yet unseen anomalies. With our novel generative approach, we transform the unsupervised task of anomaly detection to a supervised one, which is more tractable by machine learning and especially deep learning methods. DA3D surpasses the performance of state-of-the-art anomaly detection methods in a purely data-driven way, where no domain knowledge is required.