基于特征选择和SVM参数同步优化的网络入侵检测
发布时间：2019-03-11 09:37
【摘要】：为了提高网络入侵检测正确率,利用特征选择和支持向量机(SVM)参数间的相互联系,提出一种特征选择和SVM参数联同步优化的网络入侵检测算法.该算法首先将网络入侵检测正确率作为问题优化的目标函数,网络特征和SVM参数作为约束条件建立数学模型,然后通过遗传算法对数学模型进行求解,找到最优特征子集和SVM参数,最后利用KDD 1999数据集对算法性能进行测试.结果表明,相对于其他入侵检测算法,同步优化算法能够较快选择最优特征与SVM参数,有效提高了网络入侵检测正确率,加快了网络入侵检测速度.
[Abstract]:In order to improve the correct rate of network intrusion detection, a network intrusion detection algorithm based on feature selection and synchronous optimization of SVM parameters is proposed by using the relationship between feature selection and (SVM) parameters of support vector machines. The algorithm first takes the correct rate of network intrusion detection as the objective function of the problem optimization, the network features and SVM parameters as the constraints to establish a mathematical model, and then uses genetic algorithm to solve the mathematical model. The optimal feature subset and SVM parameters are found. Finally, the performance of the algorithm is tested by using the KDD 1999 dataset. The results show that compared with other intrusion detection algorithms, synchronous optimization algorithm can quickly select the optimal features and SVM parameters, effectively improve the correct rate of network intrusion detection, and accelerate the network intrusion detection speed.
【作者单位】： 平顶山学院软件学院;平顶山学院计算机科学与技术学院;
【基金】：河南省科技计划重点项目资助(102102210416)
【分类号】：TP393.08
[Abstract]:In order to improve the correct rate of network intrusion detection, a network intrusion detection algorithm based on feature selection and synchronous optimization of SVM parameters is proposed by using the relationship between feature selection and (SVM) parameters of support vector machines. The algorithm first takes the correct rate of network intrusion detection as the objective function of the problem optimization, the network features and SVM parameters as the constraints to establish a mathematical model, and then uses genetic algorithm to solve the mathematical model. The optimal feature subset and SVM parameters are found. Finally, the performance of the algorithm is tested by using the KDD 1999 dataset. The results show that compared with other intrusion detection algorithms, synchronous optimization algorithm can quickly select the optimal features and SVM parameters, effectively improve the correct rate of network intrusion detection, and accelerate the network intrusion detection speed.
【作者单位】： 平顶山学院软件学院;平顶山学院计算机科学与技术学院;
【基金】：河南省科技计划重点项目资助(102102210416)
【分类号】：TP393.08
【参考文献】
相关期刊论文 前7条
1 李辉,管晓宏,昝鑫,韩崇昭;基于支持向量机的网络入侵检测[J];计算机研究与发展;2003年06期
2 任江涛;赵少东;许盛灿;印鉴;;基于二进制PSO算法的特征选择及SVM参数同步优化[J];计算机科学;2007年06期
3 孙宁青;;基于神经网络和CFS特征选择的网络入侵检测系统[J];计算机工程与科学;2010年06期
4 唐少先;蔡文君;;基于无监督聚类混合遗传算法的入侵检测方法[J];计算机应用;2008年02期
5 周荃;王崇骏;王s，
本文编号：2438178
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2438178.html
