首页 | 官方网站   微博 | 高级检索  
     

基于连接数据分析和OSELM分类器的网络入侵检测系统
引用本文:安尼瓦尔.加马力,亚森.艾则孜,木尼拉.塔里甫.基于连接数据分析和OSELM分类器的网络入侵检测系统[J].计算机应用研究,2017,34(12).
作者姓名:安尼瓦尔.加马力  亚森.艾则孜  木尼拉.塔里甫
作者单位:新疆警察学院 信息安全工程系,新疆警察学院 信息安全工程系,新疆财经大学 计算机科学技术工程学院
基金项目:国家自然科学基金资助项目(61762086);国家社会科学基金资助项目(13CFX055); 新疆维吾尔自治区高校科研计划重点项目(XJEDU2016I052)
摘    要:针对网络入侵的实时高效检测问题,提出一种基于网络连接数据分析和在线贯序极限学习机(OSELM)分类器的网络入侵检测系统(IDS)。首先,对入侵数据库中的网络连接数据进行分析,通过特征选择算法选择出最优特征子集。然后,迭代执行交叉验证,并通过Alpha剖析来缩减样本尺寸,以此减低后续分类器的计算复杂度。最后,利用优化后的样本特征集来训练OSELM分类器,以此构建一个网络实时入侵检测系统。在NSL-KDD数据库上的实验结果表明,提出的IDS具有较高的检测率和较低的误报率,同时检测时间较短,符合实时入侵检测的要求。

关 键 词:入侵检测系统  网络连接数据  特征选择  在线贯序极限学习机  Alpha剖析
收稿时间:2016/12/16 0:00:00
修稿时间:2017/10/17 0:00:00

A Network Intrusion Detection System Based on Connection Data Analysis and OSELM Classifier
Anwar JAMAL,Yasen AIZEZI and Munila TALIFU.A Network Intrusion Detection System Based on Connection Data Analysis and OSELM Classifier[J].Application Research of Computers,2017,34(12).
Authors:Anwar JAMAL  Yasen AIZEZI and Munila TALIFU
Affiliation:Department of Information Security Engineering,Xinjiang Police College,,
Abstract:For the issues that the problem of real-time network intrusion detection, a network intrusion detection system (IDS) based on network connection data analysis and online sequential extreme learning machine (OSELM) classifier is proposed. First, the network connection data in the intrusion database is analyzed, and the optimal feature subset is selected by the feature selection algorithm. Then, the iterative implementation of cross-validation is done, and through the Alpha profile is used to reduce the sample size so as to reducing the complexity of the subsequent classifier calculation. Finally, the OSELM classifier is trained by using the optimized sample feature set to construct a real-time network intrusion detection system. Experimental results on the NSL-KDD database show that the proposed IDS has a high detection rate and a low false alarm rate, and the detection time is short, which meets the requirements of real-time intrusion detection.
Keywords:Intrusion detection system  Network connection data  Feature selection  Online sequential extreme learning machine  Alpha profile
点击此处可从《计算机应用研究》浏览原始摘要信息
点击此处可从《计算机应用研究》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号