带优先级分类约束的决策树算法

Decision Tree for Priority Classification

在与安全相关的流量管理中,对每类应用的关注度不同。针对带优先级约束的流量分类问题,设计了基于分支信息熵的决策树分类算法。实验结果表明,该算法整体准确率较高,并且各个分类的召回率与优先级约束一致。

Different attentions are paied to the applications in traffic management related to the net- work security. According to the traffic classification with restrain of priority, a decision tree algo- rithm is proposed based on the branch entropy. The experimental results demonstrate that the overall accuracy of the algorithm is high and the recall of the applications is in accordance with the priority.