面向入侵检测系统的EBP算法研究
发布时间：2018-11-08 15:34
【摘要】：BP算法是入侵检测系统(IDS)的一种重要检测方法,但是该算法存在检测时间长、效率低下等局限性.为了解决以上问题,通过对BP算法的深入剖析,提出了一种面向入侵检测系统的新算法——EBP(Enhanced BP)算法.EBP算法通过改进误差函数、增加可自适应放大的误差信号等方法,较好地解决了IDS的检测速度慢、检测效率低下等问题.实验结果表明,EBP算法相比经典的BP算法和MBP算法,不仅在网络训练的收敛速度上有了明显改善,而且在IDS的检测效率上分别提高了14.0%和8.1%,在误报率上分别降低了4.45%和1.5%.
[Abstract]:BP algorithm is an important detection method for intrusion detection system (IDS), but it has some limitations such as long detection time and low efficiency. In order to solve the above problems, through the thorough analysis of BP algorithm, a new algorithm for intrusion detection system-EBP (Enhanced BP) algorithm is proposed. The EBP algorithm can increase the self-adaptive amplification error signal by improving the error function, and so on. The problems of low detection speed and low detection efficiency of IDS are well solved. The experimental results show that compared with the classical BP algorithm and MBP algorithm, the EBP algorithm not only improves the convergence speed of network training obviously, but also improves the detection efficiency of IDS by 14.0% and 8.1%, respectively. The false positive rate decreased by 4.45% and 1.5% respectively.
【作者单位】： 内蒙古科技大学包头师范学院计算机系;内蒙古科技大学信息工程学院计算机系;
【基金】：国家自然科学基金项目(61163025) 内蒙古自治区自然科学基金项目(2010BS0904) 内蒙古自治区高等学校科学研究基金项目(NJ10162,NJZY14242,NJZY201) 包头市科学研究基金项目(2014S2004-3-1-26)
【分类号】：TP393.08
[Abstract]:BP algorithm is an important detection method for intrusion detection system (IDS), but it has some limitations such as long detection time and low efficiency. In order to solve the above problems, through the thorough analysis of BP algorithm, a new algorithm for intrusion detection system-EBP (Enhanced BP) algorithm is proposed. The EBP algorithm can increase the self-adaptive amplification error signal by improving the error function, and so on. The problems of low detection speed and low detection efficiency of IDS are well solved. The experimental results show that compared with the classical BP algorithm and MBP algorithm, the EBP algorithm not only improves the convergence speed of network training obviously, but also improves the detection efficiency of IDS by 14.0% and 8.1%, respectively. The false positive rate decreased by 4.45% and 1.5% respectively.
【作者单位】： 内蒙古科技大学包头师范学院计算机系;内蒙古科技大学信息工程学院计算机系;
【基金】：国家自然科学基金项目(61163025) 内蒙古自治区自然科学基金项目(2010BS0904) 内蒙古自治区高等学校科学研究基金项目(NJ10162,NJZY14242,NJZY201) 包头市科学研究基金项目(2014S2004-3-1-26)
【分类号】：TP393.08
【参考文献】
相关期刊论文 前4条
1 周国雄;沈学杰;李琳;贺超英;;基于AdaBoost的网络入侵智能检测[J];系统仿真学报;2014年07期
2 杨雅辉;黄海珍;沈晴霓;吴中海;张英;;基于增量式GHSOM神经网络模型的入侵检测研究[J];计算机学报;2014年05期
3 程s，
本文编号：2318906
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2318906.html
