| 基于改进DS证据融合与ELM的入侵检测算法 |
| |
| 引用本文： | 李永忠,陈兴亮,于化龙.基于改进DS证据融合与ELM的入侵检测算法[J].计算机应用研究,2016,33(10). |
| |
| 作者姓名： | 李永忠 陈兴亮 于化龙 |
| |
| 作者单位： | 江苏科技大学 计算机科学与工程学院,江苏科技大学 计算机科学与工程学院,江苏科技大学 计算机科学与工程学院 |
| |
| 基金项目： | 国家自然科学61305058,江苏省自然科学：BK20130471 |
| |
| 摘 要： | 为了提高检测率，采用D-S证据融合技术融合多个ELM，能够提高整个检测系统的精确性。但是传统的D-S技术处理冲突信息源时并不理想。因此，本文采过引入证据之间的冲突强度将信息源划分成可接受冲突和不可接受冲突，给出了新的证据理论(Improved Dempster-Shafer，I-DS)，同时针对ELM随机产生隐层神经元对算法性能造成影响的缺陷做出了改进。通过实验表明，结合I-DS和改进的ELM能够更高速，更有效的判别入侵行为。
|
| 关 键 词： | 网络入侵检测 D-S证据理论 ELM |
| 收稿时间： | 9/4/2015 12:00:00 AM |
| 修稿时间： | 2016/8/20 0:00:00 |
Intrusion detection technologies based on improved evidence fusion of DS and ELM |
| |
| Li Yongzhong,ChenXingliang and YuHualong.Intrusion detection technologies based on improved evidence fusion of DS and ELM[J].Application Research of Computers,2016,33(10). |
| |
| Authors: | Li Yongzhong ChenXingliang and YuHualong |
| |
| Affiliation: | Jiangsu University of Science and Technology,School of Computer Science and Engineering,,Jiangsu University of Science and Technology,School of Computer Science and Engineering |
| |
| Abstract: | In order to improve the detection rate, D-S evidence fusion technology and ELM could improve overall detection system. However, D-S technology was not ideal when the traditional sources was conflict. Therefore,the text gave the new evidence theory (Improved Dempster- Shafer, I-DS) by dividing the source into acceptable and unacceptable conflict judging from intensity conflict of evidences, at same time,it made an improvement about the disadvantages. Because,the disadvantages affected the algorithm performance deficiencies for the randomly producing hidden neurons. Experiments show that combined with the I-DS and improved ELM can be a faster and more effective identification intrusion. |
| |
| Keywords: | Network intrusion detection D-S evidence theory ELM |
|
| 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 |
