摘要
模糊C均值聚类算法存在自身对初始化值敏感,及其目标函数是非凸的,容易陷入局部极值点而得不到最优解的问题。结合粒子群优化算法的全局快速搜索能力,改进了目标函数,提出了改进的模糊C均值聚类算法。通过理论分析及实验证明,该算法具有较好的全局最优解,有效地克服了传统模糊C聚类算法的缺点,在入侵检测中能获得满意的检测率和误报率。
Fuzzy C-means clustering algorithm is sensitive to its initialization of value,and its objective function is non-convex,easy to fall into local minimum points,while can＇t get the optimal solution.Combined with global fast-search capability of the particle swarm optimization algorithm,improved the objective function,and puts forward the improved fuzzy C-means clustering algorithm.Through theoretical analysis and experiments,show that the algorithm has better global optimal solution,overcomes the shortcomings of traditional fuzzy C-means clustering algorithm,can obtain satisfactory detection rate and false alarm rate in the intrusion detection.
出处
《计算机与数字工程》
2010年第3期88-91,共4页
Computer & Digital Engineering
关键词
入侵检测
模糊C均值算法
目标函数
粒子群优化算法
intrusion detection
fuzzy C-means algorithm
objective function
particle swarm optimization algorithm
