COMPARATIVE ASSESSMENT OF DEEPLEARNING METHODS FOR NETWORKINTRUSION DETECTION
Srikanth Yadav.M, Mandula China Pentu Saheb · Journal of Critical Reviews · 2020
An integral component of a wired or wireless network service is a network intrusion detection device (NIDS) for both external and internal assaults. NIDS tracks network-based threats such as malware assaults by Denial of Service ( DoS), ransomware spread, and device intrusions. Numerous deep learning methods for intrusion detection applications have been suggested. They equate three modes concerning accuracy and precision: Deep Autoencoders (DAE), Deep Neural Network (DNN), and Recurrent Neural Network ( RNN). The index of network intrusion is KDD, and its output is determined by the NSL-KDD datasets. In this paper, we offer a qualitative analysis of solutions for intrusion detection to be taught in greater depth. The findings were then compared with a conventional base algorithm utilizing multinomial logistic regression to evaluate if the deep learning models are operating on this collection of data more efficiently.