- 期刊
Intrusion Detection Algorithm Based on Residual Neural Network
摘要
The existing intrusion detection technology is not ideal for detecting new intrusions. To address the problem, this paper proposes a novel intrusion detection model based on residual neural networks, using the residual structure to fully mine the interdependence between network traffic features and predict the current network state. Compared with traditional neural networks, this approach can avoid model overfitting and has higher accuracy. Finally, this paper employs the NSL-KDD benchmark dataset to test the approach provided in this work and analyzes the effect of hyper-parameter on the model's performance. Experiments reveal that the method presented in this paper outperforms traditional neural networks.
延伸閱讀
- 陳順麒（2019）。On the Recurrent Neural Network Based Intrusion Detection System〔碩士論文，逢甲大學〕。華藝線上圖書館。https://doi.org/10.6341/fcu.M0366781
- Huang, H. P., Yang, F. C., Wang, M. T., & Chang, C. M. (2009). Intrusion Detection Based on Active Networks. Journal of Information Science and Engineering, 25(3), 843-859. https://doi.org/10.6688/JISE.2009.25.3.12
- Tseng, F. Y. (2006). AN APPLICATION OF NEURAL NETWORKS IN ANOMALY AND MISUSE DETECTION FOR INTRUSION DETECTION SYSTEM [master's thesis, Tatung University]. Airiti Library. https://www.airitilibrary.com/Article/Detail?DocID=U0081-0607200917240420
- Tsai, P. C. (2021). Automatic Search Algorithm for Integrated Neural Network [master's thesis, National Taiwan University]. Airiti Library. https://doi.org/10.6342/NTU202102547
- Wang, S. (2021). Research on Network Intrusion Recognition Based on an Intelligent Algorithm. International Journal of Network Security, 23(4), 712-717. https://doi.org/10.6633/IJNS.202107_23(4).17