IP流量分类算法中特征选择作用分析.PDF

36 16 2010 8 Vol.36 No.16 Computer Engineering August 2010 ·· 2010 A TP393 IP ( 621010) C4.5BayesnetNBD NBK IP Analysis of Feature Selection Effect on IP Traffic Classification Algorithms HUANG Jun-yi, WU Jing, ZHANG Hui (College of Information Engineering, Southwest University of Science and Technology, Mianyang 621010) AbstractThe current study is to use Machine Learning(ML) techniques and classify Internet traffic based on per-flow features. Since a lot flow features can be used for flow classification and there are many irrelevant and redundant features among them, feature selection plays a vital role in algorithm performance optimization. This paper uses two filter-based feature selection methods for classification algorithms such as C4.5, Bayesnet, NBD, NBK. Experimental results show the approach can improve computational performance without negative impact on classification accuracy. Key wordsfeature selection; IP traffic classification; Machine Learning(ML) 1 IP TCP/IP (WWWP2PDNSSMTPFTP ) 2 [1] 2.1 [2] (1) (2) Genome Campus( 1 000 )24 h (3) 248 [2-3] 1 1 Label Count Label Count WWW 328 092 DA