[1]何继玲,于威威.基于 M3 和 POSS 特征的网络流量分类研究[J].计算机技术与发展,2018,28(01):83-87.[doi:10.3969/ j. issn.1673-629X.2018.01.018]
　HE Ji-ling,YU Wei-wei.Research on Network Traffic Classification Based on Min-Max Module and POSS Feature[J].Computer Technology and Development,2018,28(01):83-87.[doi:10.3969/ j. issn.1673-629X.2018.01.018]
点击复制

基于 M3 和 POSS 特征的网络流量分类研究()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
28
期数:
2018年01期
页码:
83-87
栏目:
智能、算法、系统工程
出版日期:
2018-01-10

文章信息/Info

Title:
Research on Network Traffic Classification Based on Min-Max Module and POSS Feature
文章编号:
1673-629X(2018)01-0083-06
作者:
何继玲; 于威威
上海海事大学,上海 201306
Author(s):
HE Ji-ling; YU Wei-wei
Shanghai Maritime University,Shanghai 201306,China
关键词:
网络流量分类; 类别不平衡; 多目标演化子集选择算法; 最小最大模块化
Keywords:
网络流量分类; 类别不平衡; 多目标演化子集选择算法; 最小最大模块化
分类号:
TP301
DOI:
10.3969/ j. issn.1673-629X.2018.01.018
文献标志码:
A
摘要:
网络流量分类是网络研究和流量工程的重要基础,网络流量分类大致分为基于端口号、有效负载、主机行为和机器学
习等四种分类方法。 目前基于机器学习的方法成为了研究热点。 在机器学习过程中,特征选择可以实现数据维度约简,从而提高学习模型的泛化能力。 针对大规模的流量数据以及网络流量中存在的类别不平衡问题,将最小最大集成策略(min-max module,M3)和多目标演化子集选择算法(Pareto optimization for subset selection,POSS)应用到网络流量分类的特征选择过程中。 同时将该方法与其他特征选择方法以及经典的处理类别不平衡问题的方法进行对比。 实验结果表明,M3 策略在大部分情况下能获得较好的性能,并能有效处理网络流量中类别不平衡的问题,在流量分类应用中具有一定的实用性。
Abstract:
Network traffic classification is an important foundation of network research and traffic engineering. Network traffic classification an be divided into four classification methods like basis on port number,payload,host behavior or machine learning. At present,the machine earning method has become a research hotspot. In the process of machine learning,feature selection can reduce the dimensionality of data nd improve the generalization of learning model. In view of the class imbalance of existing large-scale network traffic flow data,min-max
module (M3) and Pareto optimization for subset selection (POSS) are applied to feature selection of network traffic classification. It is compared with other feature selection methods and classic methods of dealing with the problem of class imbalance. The experiment shows that the  M3 strategy can obtain better performance in most cases and can effectively deal with the problem of class imbalance in network traffic,which has showed its effectiveness in traffic classification.

相似文献/References:

[1]韩志艳,王健. 基于加权合成少数类过采样技术的故障诊断[J].计算机技术与发展,2016,26(09):43.
　HAN Zhi-yan,WANG Jian. Fault Diagnosis Method Based on Weighted Synthetic Minority Oversampling Technique[J].Computer Technology and Development,2016,26(01):43.
[2]许二戗,于化龙.基于粒子群的多标记阈值自适应极限学习机[J].计算机技术与发展,2019,29(04):47.[doi:10. 3969 / j. issn. 1673-629X. 2019. 04. 010]
　XU Er-qiang,YU Hua-long.An Extreme Learning Machine of Multi-label Threshold Adaptation Based on Particle Swarm Optimization[J].Computer Technology and Development,2019,29(01):47.[doi:10. 3969 / j. issn. 1673-629X. 2019. 04. 010]
[3]徐 魁,海 洋,许艺凡,等.基于伽马内核与加权 K 近邻的流量分类算法[J].计算机技术与发展,2023,33(02):214.[doi:10. 3969 / j. issn. 1673-629X. 2023. 02. 032]
　XU Kui,HAI Yang,XU Yi-fan,et al.Traffic Classification Algorithm Based on Gamma Kernel and Weighted K-Nearest Neighbors[J].Computer Technology and Development,2023,33(01):214.[doi:10. 3969 / j. issn. 1673-629X. 2023. 02. 032]

更新日期/Last Update: 2018-03-13