Performance Analysis of Machine Learning Algorithms in Intrusion Detection Systems

Abstract

With the developing technology, the need for the dissemination and protection of information is becoming increasingly important. Recently, attacks on information systems have increased significantly. In addition to the rise in the number of attacks, attacks of different types pose a great threat to systems. As a result of these attacks, institutions and users suffer serious damages. At this point, Intrusion Detection Systems (IDS) have a very important position. The pre-detection of these attacks on the systems and the preparation of the necessary reports can reduce the impact of the threats that may be encountered in the future. Recent studies are carried out so as to increase the performance of IDS. In this paper, classification was made using NSL-KDD dataset and SVM, KNN, Bayesnet, NavieBayes, J48 and Random Forest algorithms, and it was aimed to compare performance of these classifications by using WEKA. Consequently, it has been reached that the KNN algorithm had the best performance with an accuracy rate of 98.1237 %. In addition, the effect of increasing the number of folds and neighborhoods on the classification result has been examined comparatively.

Keywords

• makine öğrenimi
• ids
• knn
• machine learning
• ids
• knn