Anomaly Intrusion Detection System Based On RNN-LSTM For Cyber-Attack Classification
Amrita Bhatnagar, Arun Giri, Aditi Sharma · 2024
Cybersecurity has grown in importance in many enterprises in the current generation. The network may see the emergence of novel and varied forms of cyberattacks. Thus, a system that can identify new kinds of attacks is required. This study suggests a method that can identify various kinds of attacks. The authors of this research use recurrent neural networks (RNNs) to present a Deep Learning (DL) based NIDS and explore modeling techniques for deep learning-based intrusion detection systems. The suggested work uses machine learning techniques, specifically RNN, to develop an intelligent system. PCA and RNNLSTM are combined to create a novel algorithm. The NSL-KDD dataset is used for the experiment. For feature extraction PCA technique is used and for classification, RNN-LSTM is used which shows better accuracy and detection rate than the previously developed models using RNN. The suggested model works on the binary classification.