A network traffic classification method using support vector machine with feature weighted-degree

摘要

Currently, the network traffic classification has two important problems, which are low accuracy and high computation complexity. In order to solve these problems, a novel network traffic classification method using support vector machine with feature weighted-degree (FWD-SVM) is proposed in this study. Our method can efficiently reduce the influence on the sample distribution, relative properties, and redundancy. Through reducing the training time of traffic classification machine and the predicting time of unknown samples, our method speeds up computation performance. Using support vector machine with feature weighted-degree, our method improves the stability and the accuracy of classification. The experimental results demonstrate that the proposed method not only can greatly reduce the computation complexity, but also has higher classified accuracy.