Imbalanced Data Gravitation Classification Model Using For Internet Video Traffic Identification

摘要

In the last decade, the increasing video traffic brought big challenges for Internet management. For example, illegal videos did serious harms to Internet users. Therefore, identifying different video traffic has become an urgent issue. Unfortunately, there are few researchers concern this problem. In this paper, we try to build an accurate traffic identification model using imbalanced data gravitation classification model (IDGC). We first collected a set of video traffic data from a real network. Then a special type of feature, byte code distribution feature is extracted. Finally, IDGC is used for identification. We conduct comparing experiments between IDGC and six imbalanced learning algorithms on our data. The experimental results show the competitive performances of IDGC for video traffic identification.