For many years, numerous supervised machine learning models for anomaly detection have achieved state-of-the-art performance. In this paper, we present a novel deep Q-network architecture and a custom ...