论文部分内容阅读
将自适应小波网络模型与免疫进化算法有机结合,提出基于免疫自适应小波网络的入侵检测模型及学习算法.该模型不仅减少经典神经网络在确定参数和结构时的盲目性,而且减轻对参数初始化敏感的现象.仿真对比实验结果表明,用本文算法获得的检测率比神经网络、小波网络、免疫进化的方法都要高,收敛速度更快,同时可通过控制门限值来约束和平衡漏报率和误报率之间的关系.
The adaptive wavelet network model and immune evolutionary algorithm are combined organically to propose an intrusion detection model and learning algorithm based on immune adaptive wavelet network.This model not only reduces the blindness of classical neural network in determining parameters and structure but also reduces the initialization of parameters The results of simulation experiments show that the detection rate obtained by this algorithm is higher than those of neural networks, wavelet networks and immune evolution, and the convergence speed is faster. At the same time, the threshold can be used to control and balance the omission The relationship between rate and false alarm rate.