摘要
信息系统的安全对于个人、企业甚至国家都至关重要,因此,对信息系统安全的评估也极为严格。以军工单位的信息系统为例,标准过于冗余,且样本数量较少,使得神经网络难以达到预期的评估效果。针对这些问题,建立了“信息增益比+SGAN”的分类评估模型,通过计算信息增益比得到各项指标对信息系统安全划分的重要度,减少无关紧要的冗余指标;用重要指标量化后的样本训练SGAN,得到一个强有力的分类判别网络,来对信息系统的安全进行评估分类,同时输出样本“真伪概率”和“下类似比系数”作为专家是否要人为对该系统进行评估的参考标准。
The security of information system is very important for individuals,enterprises and even countries,so the evaluation of information system is very strict.Taking the classified information system of military industrial units as an example,its evaluation criteria are various,and the small number of samples makes it difficult for the neural network to achieve the expected evaluation effect.To solve these problems,this paper establishes a classification evaluation model of"Information Gain Ratio+SGAN".By calculating the information gain ratio,we can get the importance of each standard to the security classification of information system,and then reduce irrelevant redundant standards.Then,SGAN is trained by important standards,for obtaining a powerful discriminant/classification network to evaluate and classify information system security.At the same time,the probability of"true or false"and"lower similarity ratio coefficient"are output as the reference criteria for experts to evaluate the system artificially.
作者
李健
薛惠锋
张文涛
LI Jian;XUE Hui-feng;ZHANG Wen-tao(China Aerospace Academy of Systems Science and Engineering,Beijing 100037,China)
出处
《火力与指挥控制》
CSCD
北大核心
2021年第3期131-137,共7页
Fire Control & Command Control
关键词
信息系统安全评估
信息增益比
SGAN
下类似比系数
information system security evaluation
information gain ratio
SGAN
lower similarity ratio coefficient
