Double-Adversarial Activation Anomaly Detection: Adversarial Autoencoders are Anomaly Generators. (arXiv:2101.04645v4 [cs.LG] UPDATED)

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 …

adversarial anomaly detection detection lg