Anomaly detection algorithm based on deep autoencoder ensembles

Yunpeng Kang; Haotong Zhang; Wenhao Huang

doi:10.1117/12.2627018

11 April 2022 Anomaly detection algorithm based on deep autoencoder ensembles

Yunpeng Kang, Haotong Zhang, Wenhao Huang

Proceedings Volume 12158, International Conference on Computer Vision and Pattern Analysis (ICCPA 2021); 121580P (2022) https://doi.org/10.1117/12.2627018
Event: 2021 International Conference on Computer Vision and Pattern Analysis, 2021, Guangzhou, China

Abstract

Autoencoder and deep autoencoder have been widely used for dimensionality reduction and anomaly detection. The ensemble learning method based on autoencoders further improves the accuracy of anomaly detection. However, neural networks are easy to overfit, and the current ensemble methods based on autoencoders cannot effectively make autoencoders diversified to avoid overfitting problems. For this reason, this paper proposes an ensemble method of autoencoders. The algorithm builds a cascaded model of deep autoencoders, and resample the training set of the next neural network by the anomaly detection results of the previous neural network, thereby improving the accuracy of the overall model. Experimental results show that the accuracy of the model is significantly improved compared to the current mainstream anomaly detection algorithms.

Citation Download Citation

Yunpeng Kang, Haotong Zhang, and Wenhao Huang "Anomaly detection algorithm based on deep autoencoder ensembles", Proc. SPIE 12158, International Conference on Computer Vision and Pattern Analysis (ICCPA 2021), 121580P (11 April 2022); https://doi.org/10.1117/12.2627018

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available