摘要
针对传统信息安全风险评估方法的单一性和主观性,提出了新的基于RBF模糊神经网络的信息安全风险评估方法。用模糊集合来模糊化影响评估的因素,构造网络的输入输出,用模糊规则来模拟因素之间的关系,采用增量型模糊神经网络训练方法和批处理型模糊神经网络训练方法相结合的方法来训练网络,并对从模糊规则导出的风险等级去模糊化,得到信息系统的风险指数。搭建了该RBF模糊神经网络结构,并对网络进行了学习和训练,同时与BP神经网络做了对比实验,结果表明,该方法能对信息系统的安全性做出准确的评估。
For the traditional method of information security risk assessment is single and subjective, a new information security risk assessment method called fuzzy neural network based on RBF is proposed. The factors that impact assessment are fuzzyed by fuzzy set, so the network＇ s input and output is constructed. The relationship between factors is simulated by fuzzy rules, the training method that include incremental neural network training method and batch method of fuzzy neural network are used. At last the risk index of infor- mation system is got by fuzzy rules derived from the level of risk to the fuzzy. The RBF fuzzy neural network is built, and the network is learnt and trained, and compared to the BP neural network, the result showed that the RBF fuzzy neural network is more effectivity.
出处
《计算机工程与设计》
CSCD
北大核心
2011年第6期2113-2115,2128,共4页
Computer Engineering and Design
基金
国家自然科学基金项目(60940032
61073034)
国家"十一五"科技支撑计划重大基金项目(2006BAK01A07)
国家"十一五"科技支撑计划重点基金项目(2006BAC18B06)
