Fusion of SVM and BMPM for anomaly detection in computer networks
Yuan Liu · Jisuanji gongcheng yu sheji · 2011
To solve the problems of traditional network anomaly detection systems,such as high false positve rate,a distributed anomaly detection model and the fusion method are proposed based on the fusion of SVM and BMPM,using BP artificial network.As SVM and BMPM algorithm has a distinct detection accuracy on the same network attack,ulitlizing the advantages of the two algorithms based on the BP artificial network to build a fusion anomaly detection model.According to the experiments on the KDD99 data set,it shows that the proposed model and method can obviously reduce the false positve rate,and simultaneously improve the detection rate.