基于智能优化算法选择特征的网络入侵检测  

Network intrusion detection based on selection feature of intelligent optimization algorithm

作　　者：赵悦品[1,2] 孙洁丽[1] 

机构地区：[1]河北经贸大学,河北石家庄050061 [2]河北交通职业技术学院,河北石家庄050091

出　　处：《现代电子技术》2016年第23期86-89,共4页Modern Electronics Technique

基　　金：河北省科技厅科研课题(042135136)

摘　　要：为了改善网络入侵检测的效果,提出一种智能优化算法选择特征的网络入侵检测模型。首先采用智能优化算法对网络入侵特征进行选择,得到对检测结果有重要贡献的特征,去除无效特征;然后采用支持向量机建立入侵检测分类器,最后采用KDD99数据集对模型性能进行分析。结果表明,该模型提高了网络入侵检测的准确率,而且检测速度可以满足网络安全实际应用的要求。In order to improve the effect of network intrusion detection, a network intrusion detection model based on selec- tion feature of intelligent optimization algorithm is proposed. The intelligent optimization algorithm is used to select the network intrusion features to obtain the important contribution feature for the detection result, and remove the invalid features. The sup- port vector machine is employed to establish the classifier of intrusion detection. The KDD99 dataset is adopted to analyze the model performance. The results show that the model can improve the accuracy of network intrusion detection, and its detection speed can meet the requirement of network security practical application.

关 键 词：智能优化算法 网络入侵检测 支持向量机 入侵行为 特征选择 

分 类 号：TN915.08[电子电信—通信与信息系统] TP391[电子电信—信息与通信工程]