Computer Engineering and Applications ›› 2015, Vol. 51 ›› Issue (8): 96-101.

Previous Articles     Next Articles

Network traffic clustering algorithm based on quick solution of GMM

DANG Xiaochao1,2, MAO Pengxin1, HAO Zhanjun1,2   

  1. 1.College of Computer Science and Engineering, Northwest Normal University, Lanzhou 730070, China
    2.Gansu Province Internet of Things Engineering Research Center, Lanzhou 730070, China
  • Online:2015-04-15 Published:2015-04-29

基于快速求解高斯混合模型的流量聚类算法

党小超1,2,毛鹏鑫1,郝占军1,2   

  1. 1.西北师范大学 计算机科学与工程学院,兰州 730070
    2.甘肃省物联网工程研究中心,兰州 730070

Abstract: Based on the cluster algorithm may make classification on multiple attributes , this paper proposes a clustering algorithm based on quick solution of GMM to study the classification of network traffic and achieve a better clustering effect. It is shown that it is more appropriate on traffic clustering than other algorithm. The simulation results with matlab indicate that this method is of excellent clustering precision and after the initial clustering center of the EM algorithm, it has a better accuracy of cost estimation to solve GMM, and effectively raises the convergence speed of the EM algorithm.

Key words: K-Means algorithm, parameters initialization, Gaussian Mixture Model(GMM), traffic clustering

摘要: 基于聚类算法可以对多个属性聚类的特点,提出一种基于快速求解高斯混合模型的聚类算法,用于研究网络流量的分类,使其达到更佳的聚类效果。通过与其他算法比较,讨论了该种方法在流量聚类中的适用性。仿真结果表明,该方法聚类精度高,经过初始聚类中心后的EM算法用于求解GMM有较高的估算准确性,有效地提高了EM算法的收敛速度。

关键词: K-Means算法, 参数初始化, 高斯混合模型, 流量聚类