基于MLP和Elman混合神经网络的入侵检测系统计算机科学与技术专业论文.docxVIP

哈尔滨工业大学工学硕士学位论文 哈尔滨工业大学工学硕士学位论文 - - II - Abstract With the rapid development of internet, the security of internet has become a vital problem. In order to prevent hacks from intruding internet systems, intrusion detection becomes one of most heat subjects. Intrusion detection systems detect and prohibit the abnormal actions. By building profiles of authorized computer users, the neural network can be trained to classify the incoming computer traffic into authorized traffic or not authorized traffic. Current research in the area of Intrusion Detection based Neural Networks shows encouraging results. The Intrusion Detection based Neural Networks can improve the detecting efficiency and strengthen the ability of self study. The MLP(Multlayer Perception) neural network is used widely in intrusion detection systems. But MLP neural network can only detect events individually, cannot restore the memory of past events. The model cannot identify a large consecutive stream of events. Consequently, the hybrid MLP/Elman neural network model that can be employed for both anomaly detection and misuse detection is presented in this paper. The IDSs using the hybrid neural network can detect temporally dispersed and possibly collaborative attacks effectively because of its memory of past events. The MLP network is employed as a real-time pattern classification and the Elman network is employed to restore the memory of past events. The outputs of MLP network is the inputs of the Elman network. The IDSs using the hybrid neural network are evaluated against the intrusion detection evaluation data sponsored by U.S. Defense Advanced Research Projects Agency (DARPA). Experimental results are presented in ROC curves. Experiment shows IDSs using this hybrid neural network improve the detection rate and decrease false positive rate effectively. Key words intrusion detection; abnomaly detection; hybrid network; MLP network; Elman network - - III - 目 录 摘 要 I HYPERLINK \l _bookmark0 Abstra