Analysis of Data Mining-Based Approach for Intrusion Detection System

Abstract

Due to the widespread usage of computers nowadays, there is also a greater likelihood of cyberattacks on systems or networks than previously thought. Intrusion Detection System (IDS) can be a successful method to identify the attacks in order to lessen intrusion attacks. IDS combined with data mining techniques can be an effective approach for quickly and precisely identifying damaging attacks performed by intruders and logging or blocking these attempts as necessary. IDS is a data-based process, thus by applying data mining techniques, it becomes more effective at identifying anomalous actions by drawing on its prior knowledge. In this study, we use the IDS and data mining tools to look for systemic anomalies. In the paper that follows, we demonstrate how to create a secure network by using data mining classification algorithms for intrusion detection. The machine learning classifiers Random Tree, Naive Bayes, J48, and Random Forest are used for classification. Compare the strategies based on parameters including data accuracy, detecting useful patterns, and collecting useful information.