LEAN-DMKDE: Quantum Latent Density Estimation for Anomaly Detection (Student Abstract)

This paper presents an anomaly detection model that combines the strong statistical foundation of density-estimation-based methods with representation-learning ability deep-learning models. The method autoencoder, learns a low-dimensional representation data, density-estimation based on density matrices in end-to-end architecture can be trained using gradient-based optimization techniques. A systematic experimental evaluation was performed different benchmark datasets. results show is able to outperform other state-of-the-art methods.