Abstract—With the proliferation of internet users around the world, it is becoming imperative to make communications safer than before. A network intrusion detection system is pivotal for network security because it enables us to detect and respond to malicious traffics. There are several ways and available tools to detect attacks in a computer network but machine learning techniques are one of the most efficient methods to detect abnormal traffics precisely and accurately. In this work, a method has been demonstrated to classify if incoming network traffic is normal or anomalous using machine learning techniques. Several classifiers have been evaluated based on the NSL-KDD dataset. Experiments were conducted with k-nearest neighbor, decision tree, naȉve Bayes, logistic regression, random forest, and their ensemble approach. A basic feature selection strategy has been applied to reduce the calculation time complexity and dataset’s dimension. The highest accuracy obtained 99.5% with a 0.6% false alarm rate.
 

Index Terms—intrusion detection system, machine learning, cyber security, inductive learning

Cite: Md. Badiuzzaman Pranto, Md. Hasibul Alam Ratul, Md. Mahidur Rahman, Ishrat Jahan Diya, and Zunayeed-Bin Zahir, "Performance of Machine Learning Techniques in Anomaly Detection with Basic Feature Selection Strategy - A Network Intrusion Detection System," Journal of Advances in Information Technology, Vol. 13, No. 1, pp. 36-44, February 2022. 

Copyright © 2022 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.