A Lightweight Adaptive Intrusion Detection System
Guanzhong Dai · Microelectronics & Computer · 2009
Intrusion detection system is a critical part of information security.This paper proposes a lightweight intelligent IDS model.The data mining technique and neural networks are combined in this model.The fuzzy-neural-network(FNN)model is constructed by neural network.Data mining technique is used to mining attack features in dataset.FNN model is trained by these features.The fuzzy logic has the characteristic to provides some flexibility to the uncertain problem of intrusion detection and allows much greater complexity for IDS,The neural network has the characteristic to learn by itself.This IDS model takes advantages of these characteristics,highly reduce the constructing time of IDS model and improve the accuracy of detection.This IDS model show good performance in KDD 1999 dataset detection and can be applied to real-time intrusion detection enviroments.