AI-Driven Network Intrusion Detection System Using Machine Learning

Abstract: The Network Intrusion Detection technique is an essential component of cybersecurity for detecting malicious activities in computer networks. However, signature-based system may have limitations in detecting some new or modified attacks. In this paper, an AI-based Network Intrusion Detection System has been developed that applies supervised machine learning techniques for classifying network traffic. It performs data preprocessing and feature preparation and applies Decision Tree, Random Forest, K-Nearest Neighbors (KNN), Extra Trees, and XGBoost algorithms to the CICIDS2017, CSE-CIC-IDS2018, NSL-KDD, and UNSW-NB15 datasets. The developed models are embedded in a web application that allows uploading of CSV files, traffic prediction, attack detection, and dashboard- based analysis of results.

Keywords: Network Intrusion Detection System, Machine Learning, Network Security, Cybersecurity, XGBoost, Random Forest, CICIDS2017, Network Traffic Classification.