改进蚁群算法优化SVM参数的网络入侵检测模型研究收

基于支持向量机（support vector machine, SVM)的网络入侵检测模型泛化能力与其参数选取密切相关，因此 SVM参数优化是一个难题。为进一步提高网络入侵检测率，提出一种改进蚁群优化SVM参数算法（modified ant colony optimization algorithm-support vector machine, MACO-SVM)的网络入侵检测模型。首先采用蚁群搜索路径节点代表支持向量机参数,将网络入侵检测率做为目标函数，然后通过蚁群算法的全局寻优能力和反馈机制寻找最优 参数，并对蚂蚁进行高斯变异，克服蚁群陷入局部极值,最后将最优路径上的节点连接起来得到SVM的最优参数， 建立最优网络入侵检测模型。采用KDD99数据集对模型进行仿真实验,仿真结果表明，MACO-SVM不仅提高了网络入侵的检测效率，而且获得了更高的检测率。

Network intrusion detection model based on MACO-SVM

Generalization ability of network intrusion detection model based on support vector machine ( support vector machine ,SVM) and its parameter selection are closely related. In order to solve the SVM parameter optimization problem, aiming at this problem, in order to solve parameters optimization problem for support vector machine, a novel network intrusion detection model is proposed in this paper based on support vector machine which parameters are optimized by modified ant colony optimization algorithm. Firstly, the node of ant colony search path represents the parameters of support vector machine, network intrusion detection rate is taken as the goal function, and then global optimization and feedback mechanism of ant colony optimization algorithm is used to find the optimal path, and gauss mutation is introduced to overcome local minima, and the nodes of the optimal path are connected to form the optimal parameters of support vector machine and establish the optimal network intrusion detection model, and the simulation experiments are carried out on the KDD99 dataset. The simulation results show that the proposed model not only accelerates network intrusion detection rate, but also improves intrusion detection rate compared with reference models.