Performance analysis of intrusion detection systems using machine learning methods

With the developing technology, increasing cyber attacks and detection of these attacks are among the topics that are frequently studied today. Intrusion detection systems are methods used to detect and reduce the effects of attacks on computer networks. In this study, the types of attacks that can be done on computer networks are introduced. Performance analysis of machine learning methods were performed on the identification of different intrusion detection systems on a sample dataset, and separation of normal network flow data and attack data. For this purpose, in this thesis study, classification process was performed by using K Nearest Neighbor, Principal Component Analysis, Decision Trees and Artificial Neural Networks methods, and performance analyzes were performed. The highest accuracy and highest performance on the dataset used was achieved by using the Decision Tree method. Keywords: Intrusion detection systems (IDS), machine learning (ML), decision trees