2020/02/07 by Ziyi Yang, Teng Zhang, Yang, Ziyi +5 · 4 citations
Computer Science · Mathematics · #Algorithm #Anomaly (physics) #Anomaly Detection Techniques and Applications #Anomaly detection #Artificial intelligence #Computer science #Convex hull #Deep learning #Digital Media Forensic Detection #Encoding (memory) #FOS: Computer and information sciences #Focus (optics) #Generative grammar #MNIST database #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Mathematics #Network Security and Intrusion Detection #Pattern recognition (psychology) #Property (philosophy) #Regular polygon #cs.LG #stat.ML
paper · pdf · doi:10.48550/arxiv.2002.02669
published in arXiv (Cornell University) (Cornell University)
arxiv created 2020/02/07 · openalex publication_date 2020/02/07 · arxiv updated 2020/02/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/08
In this paper, we present a memory-augmented algorithm for anomaly detection. Classical anomaly detection algorithms focus on learning to model and generate normal data, but typically guarantees for detecting anomalous data are weak. The proposed Memory Augmented Generative Adversarial Networks (MEMGAN) interacts with a memory module for both the encoding and generation processes. Our algorithm is such that most of the encoded normal data are inside the convex hull of the memory units, while the abnormal data are isolated outside. Such a remarkable property leads to good (resp. poor) reconstruction for normal (resp. abnormal) data and therefore provides a strong guarantee for anomaly detection. Decoded memory units in MEMGAN are more interpretable and disentangled than previous methods, which further demonstrates the effectiveness of the memory mechanism. Experimental results on twenty anomaly detection datasets of CIFAR-10 and MNIST show that MEMGAN demonstrates significant improvements over previous anomaly detection methods.