Machine Learning-Based Classification Approach for Network Intrusion Detection System

Today's digital world is constantly under threat from hackers and harmful programs, making it essential to have strong systems in place to detect and stop these attacks. This paper contributes to the ongoing effort to enhance network intrusion detection system (NIDS) by exploring the application of various supervised machine learning models. The goal is to identify the best-performing machine learning model for NIDS. Specifically, the study examines the efficacy of Logistic Regression (LR), Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Random Forest (RF), and XGBoost models in detecting network intrusions. The study utilized 47,736 records from a publicly accessible source. RF and XGBoost have the highest F1 scores and accuracy rates among the models tested, with both achieving an F1 score of 99.4% and an accuracy rate of 99.3%, while LR had the lowest performance among the models deployed.