Research on network intrusion detection methods is one of the important aspects in the field of cybersecurity. In this paper, we propose a network intrusion detection method based on stacked contractive autoencoder and gated convolution to address the challenge of handling high-dimensional network data in the field of intrusion detection. This method utilizes stacked contractive autoencoder to achieve data dimensionality reduction and employs gated convolution algorithm to establish a trainable dynamic feature selection mechanism, allowing for better representation of important features. Finally, the extracted features are fed into a classifier for malicious attack detection. The proposed method was trained and tested using two intrusion detection datasets, namely NSL-KDD and UNSW-NB15, achieving overall accuracies of 89.4% and 97.9% respectively.
Network Intrusion Detection Method Based on Stacked Contractive Auto-Encoder and Gated Convolution
2023-10-11
2788977 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
BLURRED IMAGE REGION DETECTION BASED ON STACKED AUTO-ENCODER
British Library Conference Proceedings | 2018
|Exemplar-based Cascaded Stacked Auto-Encoder Networks for robust face alignment
British Library Online Contents | 2018
|Fault diagnosis system and method combining convolution auto-encoder and logistic regression
Europäisches Patentamt | 2020
|Traffic anomaly detection method based on graph convolutional neural network auto-encoder
Europäisches Patentamt | 2023
|