关键词: 网络流量, 包容性检验, 支持向量机, 组合预测

Abstract: Model selection is a key problem for combination model of network traffic, and in order to improve the forecasting accuracy of network traffic, this paper proposes a network flow combination model based on encompassing test and Support Vector Machine. A lot of single models are used to forecast the network traffic, and the merits of the model are defined by mean square error of the forecasting results, and then the appropriate single model is selected by encompassing test, and the single model prediction results are combined by Support Vector Machine to get the final forecasting result of network traffic, and the performance of model is tested by the simulation experiment. The simulation results show that the proposed model can reduce the forecasting error and has improved the forecasting accuracy of network traffic.

Key words: network traffic, encompassing test, Support Vector Machine（SVM）, combination forecast