基于自适应的神经模糊推理的医疗大数据风险访问控制研究  

Research on risk access control of medical big data based on adaptive neuro fuzzy reasoning

作　　者：于国庆 沈飞 YU Guoqing;SHEN Fei(Personnel Examination and Labor Ability Appraisal Center of Weifang,Weifang Shandong 261041,China;Human Resources and Social Security Department of Shandong Province,Jinan 250014,China)

机构地区：[1]潍坊市人事考试和劳动能力鉴定中心,山东潍坊261041 [2]山东省人力资源和社会保障厅,济南250014

出　　处：《自动化与仪器仪表》2023年第1期115-120,共6页Automation & Instrumentation

基　　金：《劳动能力鉴定业务管理信息系统研发及应用》(2017RKX09)。

摘　　要：针对传统风险预测方法预测精度低,导致医疗大数据风险访问控制效果不佳的问题,提出构建一个基于自适应神经模糊理论的风险轻量化模型。首先,对BP神经网络的基本原理进行具体分析;然后在BP神经网络的基础上,结合模糊理论知识和T-S模型特性,构建一个基于T-S的模糊神经网络模型;最后通过此模型对访问风险进行量化处理,并根据访问控制策略判断是否授予访问权限。仿真结果证明,构建的模型预测结果与实际输出结果误差均值小于le-5;在非法用户的比例小于15%时,基于自适应神经模糊理论的风险轻量化模型的精确率和召回率较高。由此说明,该模型在医疗大数据风险访问控制中具有可行性。In view of the problem that the low prediction accuracy of the traditional risk prediction method causes the poor medical big data risk access control effect,a risk lightweight model based on the adaptive neurofuzzy theory is proposed.Firstly,the basic principles of BP neural network are analyzed specifically;then a fuzzy neural network model based on T-S is constructed based on BP neural network by combining with fuzzy theoretical knowledge and T-S model characteristics.Finally,the access risk is quantified and determines whether access permission is granted according to the access control strategy.Simulation results show that the mean error of the constructed model prediction result and the actual output is less than le-5;when the proportion of illegal users is less than 15%,the risk lightweight model is less than 15%.This shows that the proposed model is feasible in medical big data risk access control.

关 键 词：医疗大数据 神经网络 模糊推理 风险访问 T-S模型 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]