Hierarchical Gaussian mixture modeling improves unified anomaly detection in NF-based methods.
MACE proposes a novel anomaly detection method that efficiently handles diverse normal patterns with a unified model in the frequency domain.