ACO和SVM选择加权特征的网络攻击监测方法

发布时间：2018-01-14 15:22

本文关键词：ACO和SVM选择加权特征的网络攻击监测方法　出处：《科技通报》2015年10期 　论文类型：期刊论文

【摘要】：特征选择和分类器设计是网络攻击监测的关键,为了提高网络攻击监测率,针对特征选择问题,提出一种蚁群算法选择特征和SVM特征加权相结合的网络攻击检测方法(ACO-SVM)。首先利用支持向量机的分类精度和特征子集维数加权构造了综合适应度指标,利用蚁群算法的全局寻优和多次优解搜索能力实现特征子集搜索;然后选择网络数据的关键特征,计算信息增益获得各个特征权重,并根据特征权重构建加权支持向量机的网络攻击分类器;最后设计了局部细化搜索方式,使得特征选择结果不含冗余特征的同时提高了算法的收敛性,并通过KDD1999数据集验证了算法有效性。结果表明,ACO-SVM有效降低了特征维数,提高了网络攻击检测正确率和检测速度。
[Abstract]:Feature selection and classifier design is the key of network attack monitoring, in order to improve the monitoring network attack rate, according to the feature selection problem, proposed an ant colony algorithm of feature selection and SVM feature weighting combination of network attack detection method (ACO-SVM). By using support vector machine classification accuracy and feature subset dimension weighted to construct a comprehensive fitness the index, using ant colony algorithm global optimization and multiple optimal solution search ability to achieve feature subset search; and then select the key features of network data, calculating the information gain and obtain the feature weights, and according to the characteristics of network attack classifier weights to construct weighted support vector machine; finally design the local refined search methods, the results without feature selection the redundant features and improve the convergence of the algorithm, and through the KDD1999 data sets show the effectiveness of the proposed algorithms. The results show that ACO-SV M effectively reduces the feature dimension and improves the accuracy and detection speed of network attack detection.

【作者单位】：顺德职业技术学院;
【基金】：广东省教育研究院教育研究课题(GDJY-2015-B-b058) 广东省教育科学“十二五”规则课题(2012JK304) 顺德职业技术学院2014年校级教学改革项目(2014-SZJGXM07)
【分类号】：TP393.08;TP18
【正文快照】： Internet规模日益增加,加上其开放性、无主管性和不设防性等特性,网络攻击手段复杂多变,而且攻击数量和危害程度日益严重,网络攻击监测一直是网络安全防御研究中的重点[1]。在网络入侵建模时,特征选择和分类器设计至关重要,它们的好坏将直接影响网络攻击监测性能[2]。由于原始

【参考文献】