Multi-svm-based Decision-making Combination Of Intrusion Detection

This paper has proposed a kind of Intrusion Detection System based on the combination of multiple SVM(Support Vector Machine) classifiers through the study on SVM and the multi-classifer combination technology.Support Vector Machine is a machine learning method based on statistical learning theory. Its most major characteristic is to enhance the learning machine to exude the ability as far as possible according to the Vapnik structure risk smallest principle, namely to obtain the small error by the limited training sample collection to be able to guarantee maintaining the small error to the independent test collection. Moreover, stemming from which the support vector algorithm is a raised optimized question, the partial optimal solution also is the overall situation optimal solution, this is other study algorithms does not compare. Therefore, this paper chooses the support vector machine to take the single classifier in the entire system.In our system, the technology about multi-classifier combination is a very important part of the entire intrusion detection system and it directly decides the system's final outcome.This paper has studied several kind of combination technologies(combination technology based on most voting,weighted voting combination technology based on classifier's performance and Dynamic Classifier Selection technology based on Local Accuracy), and separately applied them in the system.In simulation experiment, this paper choses partial data in KDD'99 dataset to make the experiment to examine the system's performance and the experiment result shows that the performance of intrusion detection system using combination technology of several SVM classifiers has the better examination performance compared to the sole SVM classifier.