A Comparison of Supervised Machine Learning Algorithms for Classification of Communications Network Traffic

Abstract

Automated network traffic classification is a basic requirement for managing Quality of Service in communications networks. This research compares the performance of six widely-used supervised machine learning algorithms for classifying network traffic. The evaluations were conducted for classification of five distinct network traffic classes and two feature selection techniques. Our comparative results show that the Random Forest and Decision Tree algorithms are promising classifiers for network traffic in terms of both classification accuracy and computational efficiency.