ABSTRACT
Intrusion detection (ID) is a critical component of information security, and the fundamental technology is the ability to reliably recognize diverse network threats. Network assaults are currently the most pressing issue confronting modern society. Network risks can affect any size network, from tiny to huge. For reducing and recognizing hostile attacks in networks, an ID system is essential. Machine learning (ML) and deep learning (DL) are currently being used in a variety of sectors, including information security, to develop effective ID systems. These ID systems are capable of automatically and timely detecting harmful threats. This chapter proposes a recurrent neural network (RNN) and supervised machine learning algorithms, extra tree, cat boost, random forest (RF) and gradient boosting (GB) classifiers that analyzes and predicts dangerous intrusions in the network. The analysis was demonstrated on NSLKDD datasets, and the findings were reported in terms of F1, recall, precision, accuracy, and the training time. The accuracy of the extra tree model was able to reach 99.6 percent when compared to cat boost, RF, GB, and RNN.
