Improve Flow Accuracy and Byte Accuracy in Network Traffic Classification

Most of the current network traffic classification approaches employ single classifier method with achieving lower accuracy under small training set. Different from high flow accuracy, byte accuracy, as an important metric for network traffic classification, is usually ignored by many researchers. To address these two problems, this paper proposes a novel classification algorithm. It combines ensemble learning with cost-sensitive learning, which enables the classification model to achieve high flow accuracy as well as byte accuracy. By evaluating our algorithm with the real 7-day traces collected at the edge of the campus network, the results show that it can averagely obtain flow accuracy of 94% as well as byte accuracy of 81%.