基于多层感知器网络的过度医疗防治绩效评价研究  

Research on Performance Evaluation on Excessive Medical Treatment Based on Multi-layer Perceptron Network

作　　者：刘慧云[1] 韩玉珍[2] 刘国栋 刘朗[2] 陈玉梅 张卓[1] 付士会 

机构地区：[1]哈尔滨医科大学,黑龙江哈尔滨150081 [2]哈尔滨医科大学附属第四医院,黑龙江哈尔滨150001 [3]黑龙江省卫生发展研究中心,黑龙江哈尔滨150030 [4]哈尔滨市第五医院,黑龙江哈尔滨150020

出　　处：《中国医院管理》2017年第6期27-29,共3页Chinese Hospital Management

基　　金：国家自然科学基金资助项目(71573067);黑龙江省应用技术研究与开发计划项目(2014G0353)

摘　　要：目的通过多层感知器网络的构建评价过度医疗防治的效果,为进一步完善过度医疗防治体系提供参考。方法基于多层感知器网络,构建了具有3层神经网络、9个隐含层节点数的过度医疗绩效评价模型。结果多层感知器网络的训练结果与专家评价(期望结果)两者评价是完全一致的。结论该多层神经网络及其评价指标的选取能够很好地应用于医疗服务中的过度医疗防治绩效评价。Objective To provide a-reference for improving the governance system of excessive medical treatment through multi-layer perception network construction to evaluate the effect of excessive medical treatment.Method A performance evaluation model of excessive medical treatment was set up,which contained 9 cryptic layer nodes and 3 layers.Result The simulation of MLP neural network was consistent with the expected result.Conclusion The selection of multilayer neural network and its evaluation index can be applied to evaluate the performance of preventing and rectifying excessive medical treatment.

关 键 词：过度医疗 多层感知器网络 绩效评价 

分 类 号：R197.3[医药卫生—卫生事业管理]