Abstract: For the features of fast upgrading,strong concealment,and great randomness possessed by net intrusion,a method for intrusion detection of network security based on semi-supervised learning is proposed.The Boosting is used to build the fuzzy classifier of intrusion detection.Genetic algorithm is used to improve the iterative training,and the final the intrusion detection model of network security is thus generated.The results show that this algorithm can effectively improve the performance and efficiency of intrusion detection of network security.Compared with SVM and other advanced methods for intrusion detection,this method can detect the various types of invasion with greater accuracy,better effect and higher application value.

Key words: network security, intrusion detection, semi-supervised learning, fuzzy classifier