Study of Neural Network Technologies in Intrusion Detection Systems
Yanwei Fu, Yingying Zhu, Haiyang Yu · 2009
In recent years, the network attack become more and more widespread and difficult in against. Intrusion Detection is a major focus of research in network security. This paper analyzes neural network (NN) methods being used in IDS, in which five different types of NNs are described: multilayer perceptrons (MLP), radial basis function (RBF), self-organizing feature map (SOFM), adaptive resonance theory (ART) and principal component analysis (PCA). An intrusion detection system combined with genetic algorithm (GA) and backpropagation (BP) network is presented. Finally, a discussion of the future NN technologies, which guarantee to enhance the detection efficiency of IDS is provided.