云计算中垃圾邮件识别的机器学习算法  

Machine Learning Algorithms for Spam Recognition in Cloud Computing

作　　者：李子乾 LI Zi-qian(State Grid Co.,Ltd.,Customer Service Center,Tianjin 300309 China)

机构地区：[1]国家电网有限公司客户服务中心,天津300309

出　　处：《自动化技术与应用》2020年第5期88-95,共8页Techniques of Automation and Applications

摘　　要：由于邮件数据呈现多维特点,很难准确区分垃圾邮件发送者和正常用户,本文提出了一种基于高斯混合模型(SIGMM)的垃圾邮件识别算法,该算法将机器学习应用于移动网络中垃圾邮件发送者的智能识别。通过使用来自云服务器的移动网络数据集将SIGMM与现实挖掘算法和混合FCM聚类算法进行比较。仿真结果表明:SIGMM在召回率、精确度和时间复杂度方面均优于以前的方法。Because of the multidimensional characteristics of mail data,it is difficult to distinguish spammers from normal users accurately.This paper proposes a spam recognition algorithm based on Gauss Mixture Model(SIGMM),which applies machine learning to the intelligent recognition of spammers in mobile networks.SIGMM is compared with real-world mining algorithm and hybrid FCM clustering algorithm by using mobile network data sets from cloud servers.The simulation results show that SIGMM is superior to previous methods in recall rate,accuracy and time complexity.

关 键 词：移动网络 物联网 垃圾邮件发送者 智能识别 机器学习 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]