Research Intrusion Detection Techniques from the Perspective of Machine Learning

Abstract

With the rapid development of the Internet services and the fast increasing of intrusion problems, the traditional intrusion detection methods cannot work well with the more and more complicated intrusions. So introducing machine learning into intrusion detection systems to improve the performance has become one of the major concerns in the research of intrusion detection. Intrusion detection systems were proposed to complement prevention-based security measures. In this paper, we first introduces the basic structure of the intrusion detection system, then analysis intrusion Detection Techniques Based on Machine Learning Method, including the Bayesian based method, the neural network based method, the data mining based method and the SVM based method.