Learning from imbalanced data for encrypted traffic identification problem

Identifying encrypted application trafic is an important issue for many network tasks including quality of service, fire-wall enforcement and security. One of the challenging problems of classifying encrypted application traffic is the imbalanced property of network data. Usually, the amount of unencrypted traffic is much higher than the amount of encrypted traffic. To date, the machine learning based approach for identifying encrypted traffic often solely focused on examining and improving algorithms. The techniques for addressing imbalanced data are rarely investigated. In this paper, we present a thorough analysis of the impact of various techniques for handling imbalanced data when machine learning approaches are applied to identifying encrypted traffic. The experiments are conducted on a wellknown network traffic dataset and the results showed that some techniques for addressing imbalanced data help machine learning algorithms to achieve better performance. © 2016 ACM.