基于DEA模型的网络安全非均一评价方法

针对网络安全评价问题,基于数据包络分析(DEA:Data Envelopment Analysis)原理,引入了一种处理只有输出指标的最优偏离DEA模型,建立了网络安全评价的非均一模型和方法。该方法把权重系数作为数学规划模型的变量,通过优化方法求解,而不需要事先确定评价指标的权重系数,克服了传统的网络安全评价模型和方法因需对建立的评价指标事先确定权重系数,使评价结果受到人们主观判断的影响,从而使评价结果更加客观公正。实例分析验证了模型与方法的有效性。

Non-Uniform Evaluation Method for Network Security Based on DEA Model

Because traditional evaluation models and methods on network security such as AHP(Analytic Hierarchy Process),fuzzy comprehensive assessment,grey systems methods,need weights determining in advance,and the evaluation results have to deal do with the weights selection especially involving subjective judgment on feelings,ideas and emotions.DEA(Data Envelopment Analysis) was introduced into network security evaluation which determines the weights and priority by solving the mathematical programining corresponding to each alternative network which is called DMU(Decision Making Unit) in DEA.The weights differ from one DMU to another and DEA is a variable weights and non-uniform method.Because the weights are determined in such a way that they should be most favorable to the DMU concerned,the obtained priority relations among alternative networks are relatively optimal and fair without subjectivity.The practical example shows the availability of the established model and method.