人工鱼群和K均值算法相融合的网络入侵检测  

Networks Intrusion Detection Based on K-means Clustering Algorithm and Artificial Fish Wwarm Algorithm

作　　者：袁芳芳[1] 

机构地区：[1]太原大学,山西太原030032

出　　处：《计算机仿真》2013年第9期274-277,共4页Computer Simulation

摘　　要：针对K均值算法存在的初始聚类中心敏感和易陷入局部最优等缺陷,利用人工鱼群算法全局寻优能力,提出一种人工鱼群和K均值算法相融合的网络入侵检测模型(AFSA-KCM)。首先采用抽样技术和最大最小距离算法获得一组较优的聚类中心和聚类数目,然后通过人工鱼群模拟自然界鱼群的觅食、聚群,追尾等行为,找到最优的聚类中心和聚类数目,最后利用K均值算法根据最优的聚类中心和聚类数目建立最优的入侵检测模型,并采用KDD CUP99数据集进行测试实验。实验结果表明,相对于其它入侵检测模型,AFSA-KCM不仅提高了网络入侵检测率,同时加快了网络入侵检测速度,可以为网络安全入侵检测提供有效保证。The paper proposed a network intrusion detection model （AFSA-KCM） based on artificial fish swarm algorithm and K-means algorithm.Firstly,the sampling technique and max-min distance algorithm were used to obtain the clustering center and the clustering number,and then the artificial fish swarm was used to find the optimal clustering center and number which simulates natural feeding,cluster,and rear-end behavior.Finally the optimal intrusion detection model was built by K-means algorithm according to the optimal number and centers of cluster,and the model was tested by using KDD CUP99 data.The experimental results show that,compared with other intrusion detection models,the proposed model improves the network intrusion detection rate and network intrusion detection speed,and it can provide effective guarantee for network security.

关 键 词：入侵检测 最大最小距离算法 人工鱼群算法 K均值算法 聚类中心 

分 类 号：TP393.4[自动化与计算机技术—计算机应用技术]