Improved Autoencoder Model With Memory Module for Anomaly Detection

As a commonly used model for anomaly detection, the autoencoder model for anomaly detection does not train the objective for extracted features, which is a downside of autoencoder model. In addition, it is well known that the autoencoder model has over-prominent reconstruction ability for anomalous...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE sensors journal Ročník 24; číslo 8; s. 12770 - 12781
Hlavní autoři: Huang, Wei, Liu, Zhen, Jin, Xiaohang, Xu, Jinshan, Yao, Xinwei
Médium: Journal Article
Jazyk:angličtina
Vydáno: New York IEEE 15.04.2024
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
Anomalies
Anomaly detection
Artificial neural networks
autoencoder
Collapse
Data models
Decoding
deep support vector data description (SVDD)
Errors
Feature extraction
Hyperspheres
memory
Memory modules
Modules
Reconstruction
Training
Vectors
ISSN:1530-437X, 1558-1748
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:As a commonly used model for anomaly detection, the autoencoder model for anomaly detection does not train the objective for extracted features, which is a downside of autoencoder model. In addition, it is well known that the autoencoder model has over-prominent reconstruction ability for anomalous data, leading to high false-negative rate. On the other hand, the deep support vector data description (SVDD) model first extracts features through deep neural network, and then map the extracted features into a hypersphere. However, the deep SVDD model has disadvantages such as feature information loss and feature collapse during training process, leading to a decrease in anomaly detection accuracy. To alleviate such drawbacks mentioned above, in this article, we propose a novel model, called improved autoencoder with memory module (IAEMM). Specifically, this model jointly learns deep SVDD model and autoencoder model to minimize the overall loss of deep SVDD error and reconstruction error, and add a memory module after encoder to amplify the difference of reconstruction error between normal and abnormal data. The proposed model can well identify abnormal hidden features and mitigate the problem of feature collapse. Experimental results on several datasets confirm the effectiveness and stability of our proposed method.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1530-437X
1558-1748
DOI:10.1109/JSEN.2024.3370965