Network Intrusion Detection System for Feature Extraction Based on Machine Learning Techniques

Abstract

Network Intrusion Detection (NID) examines data from the network in search of malicious activity to identify illegal access. This investigation will center on Network Intrusion Detection (NID) technology, its progression, and the strategic value it provides. The most common types of criminal activity involving computers are illegal access, theft, and denial-of-service assaults. In recent years, experts in computer security have developed many cutting-edge solutions to protect hosts and networks from these types of threats. Both the military and commercial sectors have a requirement for intrusion detection technology. This is because intrusion detection is the most essential study field for the future of network information security. The idea of pre-processing data before training algorithms is presented in this study. The data that has been preprocessed reveals that the present Random Forest model performs better than alternative ANN models. The NSL-KDD dataset model has an accuracy of 99.12% after data pre-processing and feature extraction have been performed.