Author(s)

Xiaoman Chi, Ximin Liu

Corresponding Author

Xiaoman Chi

Abstract

Compared with the traditional network intrusion detection method that uses the off-line method to train the intrusion detection model on the existing attack samples, even if the detection rate of the existing sample types is high, it is not effective for the new types of attack samples appearing in the network. Identification, such intrusion methods have problems such as slow speed and high cost of updating models, which is not conducive to detecting new types in the network[1]. This paper mainly focuses on the GHSOM neural network model and explores its incremental intrusion method.

Keywords

Incremental, Intrusion, Detection