A novel methodology for enhancing intrusion detection system

Abstract

An Intrusion Detection System (IDS) monitors network traffic for suspicious activity and alerts when such an activity is discovered. In this study, the NSL-KDD cup 99 dataset was used to evaluate anomaly detection from intruders. Intrusion Detection System, Distributed Denial of Service (DDoS), Deep Belief Network (DBN), Random Forest, Naïve Bayes, Security Attack, Machine Learning. Pre-processing and normalization processes were performed on the dataset with inadequate, noisy, or duplicate data. A hybrid K-means clustering algorithm is used to combine clusters, which are classified using Deep Belief Networks (DBNs), Random Forest and Naïve Bayes. The study analyzed the dataset based on accuracy, precision, F-score, and false alarm rate, among which the DBN showed better performance than the other two ML algorithms.