基于大数据分析的网络流量异常检测  

Network Traffic Anomaly Detection Based on Big Data Analysis

作　　者：何通 HE Tong(College of Information Technology,Zhuzhou Teachers College,Zhuzhou Hunan 412000,China)

机构地区：[1]株洲师范高等专科学校信息技术学院,湖南株洲1412000

出　　处：《信息与电脑》2023年第19期240-242,256,共4页Information & Computer

摘　　要：随着网络流量的规模和复杂性不断增加,网络安全问题日益突出。文章分析了基于大数据分析的网络流量异常检测。首先,分析大数据的技术原理和常见的异常网络流量类型。其次,选取CICIDS2017作为实验数据集,该数据集有153万条网络流量样本,包括77个特征和11种异常流量类型。最后,进一步构建基于随机森林作的异常流量检测模型,并提取出贡献度排名前10的特征。结果表明,随机森林分类模型对于网络流量异常检测具有较高的准确性和健壮性,能够及时发现网络中的异常行为。With the increasing scale and complexity of network traffic,network security issues are becoming increasingly prominent.The article analyzes network traffic anomaly detection based on big data analysis.Firstly,analyze the technical principles of big data and common types of abnormal network traffic.Secondly,CICIDS2017 was selected as the experimental dataset,which contains 1.53 million network traffic samples,including 77 features and 11 types of abnormal traffic.Finally,a model for anomaly traffic detection based on random forest operations is further constructed,and the top 10 contributing features are extracted.The results indicate that the random forest classification model has high accuracy and robustness in detecting network traffic anomalies,and can timely detect abnormal behavior in the network.

关 键 词：大数据 网络流量 异常检测 

分 类 号：TP393.06[自动化与计算机技术—计算机应用技术]