一种基于离散马尔可夫过程的诊断风险模型  

A Model of Diagnosis Risk Based on Discrete-time Markovian Process

作　　者：李利杰[1] 姚娅川[1] 王林[1] 徐增伟[1] 徐卫东[1] 赵寅[1] 

机构地区：[1]四川理工学院自动化与电子信息学院,四川自贡643000

出　　处：《四川理工学院学报（自然科学版）》2011年第3期358-360,共3页Journal of Sichuan University of Science & Engineering(Natural Science Edition)

基　　金：四川省教育厅重点项目(2010RY001)

摘　　要：文章建立了一个基于贝叶斯公式和马尔可夫链的诊断模型,并根据中国现有的医疗管理体制进行假设检验,对假设结果进行评价。在分类过程中应用贝叶斯决策,将医疗诊断简单情况下的二值分类进行研究,同时依据分类平均风险最小的原则给出了分类的决策函数,并应用贝叶斯理论和马尔可夫过程进行讨论。实验证实了在我国建立强制医疗责任保险制度的正确性与必要性。This paper established a model for the diagnosis based on Bayesian formula and Markov chain,making hypothesis testing for the model according to China current medical management system,and the results of hypothesis testing were evaluated.Bayesian decision was applied in classification process,studying the two-value classification of medical diagnosis in simple case.While the classification decision function is given according to the principle of minimum average risk classification.And discussion was given by applying the Bayesian theory and Markov processes.The experiments confirmed that the establishment of compulsory medical liability insurance system was correct and necessary in our country.

关 键 词：贝叶斯估计 马尔可夫过程 决策函数 二值分类 平均风险 医疗责任保险制度 

分 类 号：TH16[机械工程—机械制造及自动化]