入侵检测中贝叶斯分类器的改进和设计  

Improvement and design Bayesian classifier of intrusion detection

作　　者：杨尚森[1] 徐祥生[2] 王勤[1] 李阳[3] 

机构地区：[1]洛阳理工学院现代教育技术中心,河南洛阳471023 [2]安阳工学院现代教育技术中心,河南安阳455000 [3]空军指挥学院战役战术系,北京100089

出　　处：《计算机工程与设计》2008年第16期4177-4179,共3页Computer Engineering and Design

摘　　要：针对入侵检测系统在实时检测能力和自适应能力方面的不足,提出了一个改进的贝叶斯分类器,通过引入滑动窗口技术改善入侵检测的实时性。同时通过所设计的性能调节器对贝叶斯分类器中参数的动态设置,实现了入侵检测系统的自适应性。改进后的贝叶斯分类器有效地实现了入侵检测的实时性、主动性和自适应性。Aimed at the deficiency between real-time defect and auto adjust of intrusion detection system, an improved Bayesian classifier is implemented. The skipping window technology, is used to reform the reaction time of the IDS, and at the same time it can auto adjust the system parameters to reform the running performance of the IDS by using the designed performance adjuster. Improved Bayesian classifier can effectively reform the ability of the real-time, active and auto adjust of IDS.

关 键 词：网络安全 贝叶斯分类器 入侵检测 实时 自动调整 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]