INTRUSION DETECTION SYSTEM EMPLOYING A LOGIC-BASED APPROACH FOR RULE -DRIVEN ANALYSIS

Many cyberattacks have been mounting and destabilizing these systems for a long time, the network infrastructure of every organization is unsafe.The contemporary period has seen a sharp increase in internet use.Due to the widespread usage of the internet, attackers now have the opportunity to carry out harmful actions on the communication field.A system for detecting intrusions is necessary to stop these assaults (IDS).An effective technical system, intrusion detection systems (IDS) guard against system invasions.This research compares the Logical Analysis about Data (LAD) approach with other machine learning methods, such as Support vector machine learning (SVM), naive Bayes methods, random forest analysis (RF), along with Decision Tree (DT), using the NSL-KDD dataset.The benchmark dataset in the network field is NSL-KDD.Accuracy, recall, ROC-AUC curve, F1-score, G-mean, and detection time have all been used to compare the findings.The outcome clearly shows that the LAD approach has done better than other ML-based techniques and is capable of real-time intrusion detection.