Anomaly Intrusion Detection System using Hamming Network Approach

摘要: Intrusion detection is an interesting approach that could be used to improve the security of network system. IDS detects suspected patterns traffic on remaining open parts through monitoring user activities. The major problems existing models recognition new attacks, low accuracy, time and system adaptability. In this paper, evolving anomaly intrusion constructed using hamming MAXNET Neural Network for recognize attack class in traffic. result encouraging, rate 95% which relatively high. We describe another based Multilayer Perceptrons (MLP) compare results two approaches evaluate experimental demonstrate designed are promising terms accuracy computational real word detection. Training testing data obtains from Defense Advanced Research Projects Agency(DARPA) evaluation datasets.