Research On Ids Based On Support Vector Machine

With the fast development of the computer network, people's reliance on the network grows with each passing day, however, the following one is the overflowing of the attack, the network security has already become a focus that is attracting people's attention, and occupy the important position in the national security system.The intrusion detection technology has become an essential technology in the information security, it becomes the new hot spot of the network security in recent years. The high-speed development of Internet, has proposed the new challenge in intrusion detection technology. The existing invasion examination system mostly based on the rule examination, the speed is slow, the examination rate of accuracy is not high. At present, many security software has provided the partial functions of intrusion detection, but as a result of the multiplication, the intellectualization, the complication of the network attack method, there still have the disadvantage of high false positive and rate of missing report. The support vector machines method solves this kind of problem well.When the network dataset is very large, conventional Support Vector Machine(SVM) learning algorithm is remarkably slow. In contrast, the proposed algorithm based on space block and sample density is fast. It is applied in intrusion detection in this paper. The algorithm selects training samples by local sample density, to reduce the training samples and thus to improve the speed of learning. Simulation shows that the algorithm is faster than the techniques of intrusion detection based on conventional SVM while guarantee the high classification precision.This paper proposes a classification algorithm based on CS-SVM and Bagging. As every sample be defined different cost, with the means of minimizing the misclassification cost, it comes to such result that the algorithm brings the optimal classifier to ensure the precision of classification. Simulation results show that the algorithm has high precision and efficiently lowers the error ratio.