%0 Journal Article
%T Anomalous traffic classification based on maximum entropy model<br>基于最大信息熵模型的异常流量分类方法*
%A QIAN Ya-guan
%A GUAN Xiao-hui
%A WANG Bin
%A <br>钱亚冠
%A 关晓惠
%A 王滨
%J 计算机应用研究
%D 2012
%I 
%X The machine learning model based on maximum entropy principles has been applied successfully in natural language processing, such as machine translation, text auto-classification and speech recognition. This model was first used in network anomalous traffic classification with our exploration. As the maximum entropy model used binary feature function, which was fit for processing nominal feature, it adopted the discrete method based on entropy to preprocessing the training data set. It generated the final feature set by extracting features from KDD99 dataset with CFS algorithm. Finally, employed the BLVM algorithm to evaluate the parameters and got an exponential model subjected to maximum entropy constrain. The model was compared with Naive Bayes, Bayes Net, SVM and C4.5 by precision, callback and F-Measure. The results of experiment show that the maximum entropy model has better classification efficiency, especially under small size of training data set.
%K maximum entropy model
%K anomalous traffic
%K discretezation
%K feature selection
%K parameter evaluation<br>最大信息熵模型
%K 异常流量
%K 离散化
%K 特征选择
%K 参数估计
%U http://www.alljournals.cn/get_abstract_url.aspx?pcid=5B3AB970F71A803DEACDC0559115BFCF0A068CD97DD29835&cid=8240383F08CE46C8B05036380D75B607&jid=A9D9BE08CDC44144BE8B5685705D3AED&aid=F950C016BC33E90185AE53D9FDCE2892&yid=99E9153A83D4CB11&vid=771469D9D58C34FF&iid=38B194292C032A66&sid=CBC69BEA05C12902&eid=A40901B6135B333E&journal_id=1001-3695&journal_name=计算机应用研究&referenced_num=0&reference_num=21