基于支持向量机方法的网络入侵检测实验研究

网络信息不断增加和攻击手段日益复杂,给网络安全领域带来了日益严峻的挑战.为了改善网络入侵检测技术现状,提出了一种基于支持向量机和决策集合理论融合的网络入侵检测方法,通过对规则信息、攻击信息、边界信息的准确界定完成检测过程.选取了基于神经网络的入侵检测方法、基于遗传算法的入侵检测方法、基于传统支持向量机的入侵检测方法作为对比算法,在K-Cup测试数据集下展开实验研究.实验结果表明,该文提出的方法具有更高的召回率、精确率、查准率和更低的误检率,其性能明显优于其他3种方法,可应用于入侵检测领域.

Support Vector Machine Method for Network Intrusion Detection

With the increasing of network information and the increasing complexity of attack means, the network security field is facing more and more severe challenges. In order to improve the current situation of network intrusion detection technology, a method of network intrusion detection based on the fusion of support vector machine and decision set theory has been proposed. The detection process is completed by defining the rule information, attack information and boundary information accurately. The intrusion detection methods based on neural network, genetic algorithm and traditional support vector machine have been selected as comparison algorithms, and experiments been carried out under K-Cup test data set. The experimental results show that the proposed method has higher recall rate, accuracy rate, precision rate and lower false detection rate, and its performance is obviously superior to the other three methods, which can be applied in the field of intrusion detection.