基于粒子群优化的k均值算法在网络入侵检测中的应用  

Research of K-means algorithm based on particle swarm optimization in network intrusion detection

作　　者：谷保平[1] 许孝元[1] 郭红艳[2] 

机构地区：[1]广东工业大学计算机学院,广东广州510006 [2]华侨大学信息学院,福建泉州362021

出　　处：《计算机应用》2007年第6期1368-1370,共3页journal of Computer Applications

摘　　要：在k均值算法基础上,提出一种将粒子群算法与k均值算法相结合产生基于粒子群的k均值算法(PSO-k均值算法)。用KDD cup99数据集进行评估k均值算法和PSO-k算法检测性能。试验结果表明,PSO-k均值算法能够避免k均值算法固有的缺点,检测率提高和误报率下降,并且有较高的检测性能。Based on the k-means algorithm, we proposed the PSO-k-means algorithm combining the k-means with PSO. Then KDD cup 99 dataset was applied to both algorithms. The experiment indicates that the PSO-k-means algorithm can avoid the inherent shortcomings of the k-means algorithm, and has higher detection performance with detection rate rising and false alarm rate falling.

关 键 词：K均值算法 粒子群算法 PSO-k均值 入侵检测 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]