Twin Fuzzy Networks With Interpolation Consistency Regularization for Weakly-Supervised Anomaly Detection

Zhi CAO, Ye SHI, Yu-Cheng CHANG, Xin YAO, Chin-Teng LIN

Research output: Journal PublicationsJournal Article (refereed)peer-review

Abstract

Weakly-supervised anomaly detection (WSAD) has gained increasing attention due to its core idea of enhancing the performance of unsupervised anomaly detection by leveraging prior knowledge from a limited number of labeled anomalies. In this paper, we introduce a novel WSAD framework that surpasses current state-of-the-art methods in terms of accuracy, exhibits greater robustness to data uncertainty, and is more efficient in utilizing limited labeled anomalies. Our method is built upon twin fuzzy networks (TFN) that learn robust fuzzy if-then rules from a pairwise training set. TFN can extract informative prototypes of training instances, exploiting the very few labeled anomalies efficiently. A two-stage sequential training scheme, comprising fuzzy C-means clustering and interpolation consistency regularization, ensures that the fuzzy rules form a solid foundation for anomaly detection while improving TFN's generalization ability. The training process of TFN relies on closed-form optimization rather than gradient-based methods, leading to significantly faster training speeds. Comprehensive experiments conducted on numerous real-world datasets confirm the advantages of the TFN framework over existing alternatives.
Original languageEnglish
Number of pages12
JournalIEEE Transactions on Fuzzy Systems
DOIs
Publication statusE-pub ahead of print - 11 Jun 2024

Bibliographical note

Publisher Copyright:
IEEE

Keywords

  • Anomaly detection
  • Data uncertainty
  • Interpolation
  • Knowledge engineering
  • Optimization
  • Prototypes
  • Training
  • Uncertainty
  • fuzzy c-means clustering
  • interpolation consistency regularization
  • twin fuzzy networks
  • weakly-supervised anomaly detection

Fingerprint

Dive into the research topics of 'Twin Fuzzy Networks With Interpolation Consistency Regularization for Weakly-Supervised Anomaly Detection'. Together they form a unique fingerprint.

Cite this