摘要
针对数据中各个字段属性差异及其对产生入侵行为的作用度分析不足,从缓解模糊入侵检测中误差率高入手,验证其中存在的等价转换失真问题,用动态自反馈理论改造模糊聚类过程,并分析入侵数据类型及其在入侵中所起作用,建立面向混合数据的自反馈模糊聚类方法,并在此基础上构建入侵检测系统。实验表明本方法能够有效提高入侵检测引擎的检测率,降低其误报率,缓解上述问题。
In view of the differences in the attributes of each word section in the data and the deficiency in analyzing the role degree of intrusion and for the purpose of mitigating the high error rate in fuzzy intrusion detection, the problem of equivalent conversion distortion was tested and verified and fuzzy clustering process was reformed with dynamic seef feed back theory. The types of intrusion data and their roles in the intrusion were analyzed. The seef feedback fuzzy clustering method for mixed data was established. On the above basis, the intrusion detecting system was constructed. Experiments prove that this method can effectively increase the detecting rate of the intrusion detecting engine, reduce its error rate and mitigate the above problem.
出处
《辽宁工程技术大学学报（自然科学版）》
CAS
北大核心
2008年第6期881-884,共4页
Journal of Liaoning Technical University (Natural Science)
基金
广西教育厅科研基金资助项目(200705LX300)
广西财经学院科技基金资助项目(2007B15)
