Application Of Support Vector Machine In Intrusion Detection System

With the rapid development and wide application of network technologies, network security is becoming more and more important. It is a very urgent problem in intrusion detection system that how to recognize existing attacks and increasingly new attacks rapidly,exactly and effectively.Compared to traditional intrusion detection technologies, pattern recognition technology yielded encouraging effects on intrusion detection system based on machine learning. The generalization of pattern recognition can make to recognize new or unknown attacks and classification capability of classifier can improve accuracy of intrusion detection. As built on structural risk minimization and VC dimension theory of Statistical Learning, SVM is not easy to be run into local optimum and conquers curse of Dimensionality handily. And that special advantages on intrusion classification problem which has small samples and high dimension.Considering the fact that each feature of the network connection records has different effect on the detection result, a feature selected method according to the character of the decision function of support vector machine and the study of previous researchers was put forward, so as to improve the efficiency and accuracy. The primary contents of research are as follows:(1)The development course of intrusion detection system was introduced, the current research situation of intrusion detection system was expatiated and the virtues and limitations of existing and detect technical were discussed.(2)The basic theories of statistical learning theory were expatiated; When the support vector machine was expounded, the reason that support vector machine was superior to the other learning algorithms were analyzed. And the implementation methods were introduced.(3)Since different feature in the network-feature data has different degree of impact of classification accuracy, a new feature select method according to the decision function which could be represented as sum of products of weights and features was putted forward based on the study of previous researchers.Results from experiments on the KDDCUP99 dataset indicated that the method was effective and efficient.