Research of SVM with Normalization in Network Intrusion Detection

摘要：

Network intrusion is always hidden in a mass of routine data and the difference between the data and data in the same feature is very large. Normalization brings the advantages of speeding up the learning phase and avoiding numerical problems such as precision loss from arithmetic overflows. Many normalization methods are analyzed.Because Max Normalization and Min-Max Normalization has simple rules and fast speed,they are used to experiment.Min-Max Normalization overcomes boundary problem in Max Normalization. Experiments show that the SVM with normalization has obvious improved compare with the SVM without normalization in classing intrusion data of KDD99. Min-Max Normalization also has better performance in speed,accuracy of cross validation and quantity of support vectors.

作者: Shengli Xie Zhenyu Liu

作者单位: College of Electronic and Information Engineering South China University of Technology

会议类型: 国际会议

会议名称: Fourth International Conference on Impulsive and Hybrid Dynamical Systems(ICIHDS 2007)(第四届国际脉冲和混合动力系统学术会议)

会议地点: 南宁

会议语种:英文

在线出版日期: 2007-07-20（万方平台首次上网日期，不代表论文的发表时间）