基于人工神经网络的医保定点医疗机构信用等级评价模型
|
摘要
在城镇职工基本医疗保险改革的过程中,医保定点医疗机构出现了大量的问题,这些问题严重的影响了医疗保险基金的安全运行。为解决这一问题,2004后政府采用了对医保定点医疗机构进行信用等级评价的管理办法。
为实行对医保定点医疗机构进行信用等级评价的管理办法,各地政府都采用了一些技术手段对医保定点医疗机构进行信用等级评价。根据所得到的材料来看,这些信用等级评价的方法主要是“专家评价法”。这种评价方法过于依赖评估人员的经验和能力,主观性较强,结果的客观、公正性难以保证。不仅如此,这种评价方法还需要大量的资金支持,对医疗保险基金造成了浪费,这与对医保定点医疗机构进行信用等级评价的初衷相违背。
为了克服“专家评价法”的不足,本文利用人工神经网络模型对医保定点医疗机构的信用等级进行学习,并且根据学习过程中出现的问题,对人工神经网络做了改进,克服了医保定点医疗机构信用等级评价网络原有的不足。最后经过试验的分析与证明,学习所得出来的医保定点医疗机构信用等级评价网络是有效的,这个模型能被用于以后评价医保定点医疗机构进行信用等级的工作当中去。
为了再进一步加强医疗保险中心对医保定点医疗机构的管理,在本文中,以医保信息系统形成的大量数据为基础,利用LOF算法对大量数据进行挖掘,找出了医保定点医疗机构的违规行为。最后经过试验证明:使用这种方法所找出的违规行为的准确率在60%以上,可见种方法是有效地。由于这种方法的使用,大大减轻了人工核实医保定点医疗机构违规行为的负担。
In the reform process of basic medicare of town officers and workers, there are a large of problems in medical organizations appointed by medicare center,and the problems made medicare fund in danger. for solve these problems,since 2004,government start manage medical organizations by credit rank evaluation.
For carry out manage method that evaluate the credit rank of medical organizations appointed by medicare center, the governments all adopted some technique means to evaluate the credit rank of medical organizations appointed by medicare center. According the material about medicare, the methods of evaluation about credit rank is mainly "expert evaluation method". But his evaluation method is too depend on the personnel's experience and ability to impersonality.Not only such, this evaluation method still need a great deal of funds for carry out, resulted in that a large of medicare fund be wasted, this oppose the original intention of credit rank evaluation of medical organizations appointed by medicare center.
For overcoming the shortage of"expert evaluation method", this text make use of the artificial nerve network model to carry out learning evaluate the credit rank of medical organizations appointed by medicare center, and according the problem of appearing in the learning process to make an improvement to the artificial nerve network, at last,the improvement overcame the shortage of Artificial Neural Networks that evaluate credit rank of medical organizations appointed by medicare center. At last,by the analysis and proof of the experiments that has been done, credit rank evaluation networks that has been learned of medical organizations appointed by medicare center is effective,this model can be used for a later evaluation.
For made a more strict manage to the medical organizations appointed by medicare center,in paper,base on a large of date in medicare datebase,find out the rules-out behavior of medical organizations appointed by medicare center by the way that make use of LOF algorithm to dig date. Finally,the experiment that has been done prove:the exactitude rate of rules-out behavior that has been find above 60%,thus it can be seen,this method is effective.due to the method be used,the working load of medicare center's workers be lightened.
为实行对医保定点医疗机构进行信用等级评价的管理办法,各地政府都采用了一些技术手段对医保定点医疗机构进行信用等级评价。根据所得到的材料来看,这些信用等级评价的方法主要是“专家评价法”。这种评价方法过于依赖评估人员的经验和能力,主观性较强,结果的客观、公正性难以保证。不仅如此,这种评价方法还需要大量的资金支持,对医疗保险基金造成了浪费,这与对医保定点医疗机构进行信用等级评价的初衷相违背。
为了克服“专家评价法”的不足,本文利用人工神经网络模型对医保定点医疗机构的信用等级进行学习,并且根据学习过程中出现的问题,对人工神经网络做了改进,克服了医保定点医疗机构信用等级评价网络原有的不足。最后经过试验的分析与证明,学习所得出来的医保定点医疗机构信用等级评价网络是有效的,这个模型能被用于以后评价医保定点医疗机构进行信用等级的工作当中去。
为了再进一步加强医疗保险中心对医保定点医疗机构的管理,在本文中,以医保信息系统形成的大量数据为基础,利用LOF算法对大量数据进行挖掘,找出了医保定点医疗机构的违规行为。最后经过试验证明:使用这种方法所找出的违规行为的准确率在60%以上,可见种方法是有效地。由于这种方法的使用,大大减轻了人工核实医保定点医疗机构违规行为的负担。
In the reform process of basic medicare of town officers and workers, there are a large of problems in medical organizations appointed by medicare center,and the problems made medicare fund in danger. for solve these problems,since 2004,government start manage medical organizations by credit rank evaluation.
For carry out manage method that evaluate the credit rank of medical organizations appointed by medicare center, the governments all adopted some technique means to evaluate the credit rank of medical organizations appointed by medicare center. According the material about medicare, the methods of evaluation about credit rank is mainly "expert evaluation method". But his evaluation method is too depend on the personnel's experience and ability to impersonality.Not only such, this evaluation method still need a great deal of funds for carry out, resulted in that a large of medicare fund be wasted, this oppose the original intention of credit rank evaluation of medical organizations appointed by medicare center.
For overcoming the shortage of"expert evaluation method", this text make use of the artificial nerve network model to carry out learning evaluate the credit rank of medical organizations appointed by medicare center, and according the problem of appearing in the learning process to make an improvement to the artificial nerve network, at last,the improvement overcame the shortage of Artificial Neural Networks that evaluate credit rank of medical organizations appointed by medicare center. At last,by the analysis and proof of the experiments that has been done, credit rank evaluation networks that has been learned of medical organizations appointed by medicare center is effective,this model can be used for a later evaluation.
For made a more strict manage to the medical organizations appointed by medicare center,in paper,base on a large of date in medicare datebase,find out the rules-out behavior of medical organizations appointed by medicare center by the way that make use of LOF algorithm to dig date. Finally,the experiment that has been done prove:the exactitude rate of rules-out behavior that has been find above 60%,thus it can be seen,this method is effective.due to the method be used,the working load of medicare center's workers be lightened.
引文
[1]王文娟,陈岱云.中国医疗保险的现状与发展趋势分析.理论学刊,2009(10):51~55.
[2]仇雨临,孙树菡.医疗保险.中国人民大学出版社,2001.
[3]朱理平.城镇职工医疗保险基金管理使用中存在的问题及对策.现代审计与经济,2008(2):31.
[4]王鸿勇.山东省医疗保险定点医院信用等级制度对医疗保险费用控制效果的影响作用研究.博士学位论文.
[5]郭敏华.信用评级[M].北京:中国人民大学出版社,2004.
[6]李乐波,刘翰林,楼欢欣.我国医疗机构信用评价研究[J].中国医院,2006,10(1):42~47.
[7]陈希镇,曹慧珍.判别分析和spss的使用.科学技术和工程,2008,8(13):35~39.
[8] K. Fukunaga, J. M. Mantock. Nonparametric Discriminant Analysis. IEEE Transactions on Pattern Analysis and Machine Intelligence. 1983(5): 671~678.
[9] K. Fukunaga. Introduction to statistical pattern recognition. Academic Press, Boston, 2nd edition, 1990.
[10] M. Bressan, J. Vitria. Nonparametric discriminant analysis and nearest neighbor classification. Pattern Recognition Letters. 2003(12): 2743~2749
[11] Xipeng Qiu, Lide Wu. Face Recognition by Stepwise Nonparametric Margin Maximum Criterion. IEEE International Conference on Computer Vision. 2005, l(2): 1567~1572.
[12] M. Loog, R. P. W. Duin. Linear Dimensionality Reduction via a Heteroscedastic Extension of LDA. The Chernoff Criterion. IEEE Trans. Pattern Analysis and Machine Intelligence, 2004, 6(26): 732~739.
[13] Z. Li, W. Liu, D. Lin, X. Tang. Nonparametric subspace analysis for face recognition. Computer Vision and Pattern Recognition, 2005, 1(2): 961~966.
[14]施朝建,张明铭.Logistic回归模型分析析.计算机辅助工程,2005,14(13):46~50.
[15] Ulf,Nilsson,.Jan,Maluszynski..LOGIC:PROGRAMMING,AND,PROLOG..2000.
[16] West. R. C., A factor-analytic approach to bank condition[J]. Journal of Banking and Finance, 1985.
[17] Lawrence. E. L.. An analysis of default risk in mobile home credit[J]. Journal of Banking and Finance, 1992.
[18] Smith, Lawrence. Forecasting losses on a liquidating longterm loan portfolio[J]. Journal of Banking and Finance,1995.
[19]曾华军,张银奎.机器学习.机械工业出版社,2003.
[20] L..P..J..Veelenturf..Analysis,and,Applications,of,Artificial,Neural,Networks.. 1995.
[21] E. I. Altman, Giancarlo Marco, Franco Varetto.Corporate distress diagnosis: comparison using linear discriminate analysis and neural networks [J]. Journal of Banking and Finance, 1994, 18(1): 505~5291.
[22] Trippi. R. R., E. Turban. Neural networks in finance and investing. Chicago, Irwin Prof essional Publishing, 1996.
[23]陈祥光,裴旭东.人工神经网络技术及应用.中国电力出版社,2003.
[24]杨建刚.人工神经网络实用教程.浙江大学出版社,2001(2):47~58.
[25]田彦梅.从金融机构的立场看医疗机构的等级划分.国外医学,2001,18(4):173~178.
[26]朱盈权.《企业质量信用等级划分通则》国家标准颁布2009年11月1日起实施.中国质量报,2009,28(9):49.
[27]张建.《企业质量信用等级划分通则》国家标准颁布.信息技术与标准化,2009(7):8.
[28]劳动和社会保障部社会保险事业管理中心.关于开展城镇职工基本医疗保险定点医疗机构信用等级评定工作的意见.劳社险中心函,2004(58).
[29]陈元燮.建立信用评级指标体系的几个理论问题.财经问题研究,2000(8):23~27.
[30]顾基发.意见综合—怎样达成共识[J].系统工程学报,2001,16 (5):340~348.
[31]徐国祥.统计预测与决策.上海财经大学出版社,2008(3):86~101.
[32]张芳,黎玉柱.社会医疗保险中道德风险的表现与成因分析.中国卫生事业管理,2007(5):314~319.
[33]长沙市医疗保险中心.医疗待遇数据字典,2007年.
[34]夏火松.数据仓库与数据挖掘技术[M].科学出版社,2004(3):135~160.
[35]张云涛,龚玲.数据挖掘原理与技术[M].北京:电子工业出版社,2004(1):181~213.
[36] Malik Agyemang. Local Sparsity Coeficient-Based Mining of Outliers[D]. University of Windsor, 2002.
[37] Han Jiawei, KamberM. Data Mining: Concepts and Techniques[M ]. Academic Press, 2001.
[38] Hawkins. D.. Identification of Outliers[M]. London: Chapman and Hall, 1980.
[39]卢辉斌,徐刚,李段.一种基于孤立点检测的入侵检测方法[J].微机发展,2005,15(6):93~94.
[40] Kuramitsu..K../Finding.Periodic.Outliers.over.a.Monogenetic.Event.Stream[C]..P-roc..of.International.Workshop.on.Ubiquitous.Data.Management..Tokyo,.Japan:[S.n.],.2005: 97~104.
[41]杨永铭,王喆.孤立点挖掘算法研究.计算机与数字工程,2008(1):11~15.
[42] Sherhar..S.,.Lu..C.,.Zhang.P...A.unified.approach.to.detecting.spatialoutliers[J]..Geolnfomlatica,.2003.
[43] Breunig. M. M.,.Kriegel. H. P., Ng. R., et al. LOF: Identifying Density-Based Local Outliers[C]. Dallas, Texas: Proc. of ACM SIGMOD Conf, 2000: 201~235.
[44]杨风召,朱扬勇,施伯乐.IncLOF:动态环境下局部异常的增量挖掘算法[J].计算机研究与发展,2004,41(3):477~484.
[45]蒋宗礼.人工神经网络导论.高等教育出版社,2001.
[46] Shavilik. J. W.,.Mooney. R. J...Symbolic.and.neural.learning.algorithms..Maching. Learning,.1991(6): 111~114.
[47] Weiss..S.,.Kapouleas..I...An.empirical.comparison.of.pattern.recognition,.neural. nets.and machine learning classing classification methods. Proceedings of the Eleventh IJCAI, 1989.
[48] Vonder Malshurg. C. Sels-organizing of orientation sensitive cells in the striate cortex. Kybernetikl. 1973.
[49] Lippmann. R. P.. An Introduction to Computing With Neural Nets[M]. IEEE ASSP Magazine, 1999: 23~45.
[50] Cyberko. G.. Approximations by superpositions of a sigmoidal function[A]. Math Control Singnal System[C], 1989.
[51] Hecht-Nielsen. R. Theory of back propagation neural network[J]. Proc of IJCNN, 1989(1): 593~603.
[52] Kennedy..J.,.Eberhart..R..C...Particle.swarm.optimization[A]..Proc..IEEE.Int..Co-nf. on Neural Networks[c], Perth WA Australia, 1995.
[53] Eberhart..R..C.,.Kennedy..J..A..new.optimizer.using.particle/swarm.theory[A]..Pr-oc.the.Sixth.lnt..Symposium.on.Micro.machine.and.Human.Science[C]..Nagoya, Japan, 1995: 39-43.
[54]谢晓锋,张文俊,杨之廉.微粒群算法综述.控制与决策,2003(3):129~134.
[2]仇雨临,孙树菡.医疗保险.中国人民大学出版社,2001.
[3]朱理平.城镇职工医疗保险基金管理使用中存在的问题及对策.现代审计与经济,2008(2):31.
[4]王鸿勇.山东省医疗保险定点医院信用等级制度对医疗保险费用控制效果的影响作用研究.博士学位论文.
[5]郭敏华.信用评级[M].北京:中国人民大学出版社,2004.
[6]李乐波,刘翰林,楼欢欣.我国医疗机构信用评价研究[J].中国医院,2006,10(1):42~47.
[7]陈希镇,曹慧珍.判别分析和spss的使用.科学技术和工程,2008,8(13):35~39.
[8] K. Fukunaga, J. M. Mantock. Nonparametric Discriminant Analysis. IEEE Transactions on Pattern Analysis and Machine Intelligence. 1983(5): 671~678.
[9] K. Fukunaga. Introduction to statistical pattern recognition. Academic Press, Boston, 2nd edition, 1990.
[10] M. Bressan, J. Vitria. Nonparametric discriminant analysis and nearest neighbor classification. Pattern Recognition Letters. 2003(12): 2743~2749
[11] Xipeng Qiu, Lide Wu. Face Recognition by Stepwise Nonparametric Margin Maximum Criterion. IEEE International Conference on Computer Vision. 2005, l(2): 1567~1572.
[12] M. Loog, R. P. W. Duin. Linear Dimensionality Reduction via a Heteroscedastic Extension of LDA. The Chernoff Criterion. IEEE Trans. Pattern Analysis and Machine Intelligence, 2004, 6(26): 732~739.
[13] Z. Li, W. Liu, D. Lin, X. Tang. Nonparametric subspace analysis for face recognition. Computer Vision and Pattern Recognition, 2005, 1(2): 961~966.
[14]施朝建,张明铭.Logistic回归模型分析析.计算机辅助工程,2005,14(13):46~50.
[15] Ulf,Nilsson,.Jan,Maluszynski..LOGIC:PROGRAMMING,AND,PROLOG..2000.
[16] West. R. C., A factor-analytic approach to bank condition[J]. Journal of Banking and Finance, 1985.
[17] Lawrence. E. L.. An analysis of default risk in mobile home credit[J]. Journal of Banking and Finance, 1992.
[18] Smith, Lawrence. Forecasting losses on a liquidating longterm loan portfolio[J]. Journal of Banking and Finance,1995.
[19]曾华军,张银奎.机器学习.机械工业出版社,2003.
[20] L..P..J..Veelenturf..Analysis,and,Applications,of,Artificial,Neural,Networks.. 1995.
[21] E. I. Altman, Giancarlo Marco, Franco Varetto.Corporate distress diagnosis: comparison using linear discriminate analysis and neural networks [J]. Journal of Banking and Finance, 1994, 18(1): 505~5291.
[22] Trippi. R. R., E. Turban. Neural networks in finance and investing. Chicago, Irwin Prof essional Publishing, 1996.
[23]陈祥光,裴旭东.人工神经网络技术及应用.中国电力出版社,2003.
[24]杨建刚.人工神经网络实用教程.浙江大学出版社,2001(2):47~58.
[25]田彦梅.从金融机构的立场看医疗机构的等级划分.国外医学,2001,18(4):173~178.
[26]朱盈权.《企业质量信用等级划分通则》国家标准颁布2009年11月1日起实施.中国质量报,2009,28(9):49.
[27]张建.《企业质量信用等级划分通则》国家标准颁布.信息技术与标准化,2009(7):8.
[28]劳动和社会保障部社会保险事业管理中心.关于开展城镇职工基本医疗保险定点医疗机构信用等级评定工作的意见.劳社险中心函,2004(58).
[29]陈元燮.建立信用评级指标体系的几个理论问题.财经问题研究,2000(8):23~27.
[30]顾基发.意见综合—怎样达成共识[J].系统工程学报,2001,16 (5):340~348.
[31]徐国祥.统计预测与决策.上海财经大学出版社,2008(3):86~101.
[32]张芳,黎玉柱.社会医疗保险中道德风险的表现与成因分析.中国卫生事业管理,2007(5):314~319.
[33]长沙市医疗保险中心.医疗待遇数据字典,2007年.
[34]夏火松.数据仓库与数据挖掘技术[M].科学出版社,2004(3):135~160.
[35]张云涛,龚玲.数据挖掘原理与技术[M].北京:电子工业出版社,2004(1):181~213.
[36] Malik Agyemang. Local Sparsity Coeficient-Based Mining of Outliers[D]. University of Windsor, 2002.
[37] Han Jiawei, KamberM. Data Mining: Concepts and Techniques[M ]. Academic Press, 2001.
[38] Hawkins. D.. Identification of Outliers[M]. London: Chapman and Hall, 1980.
[39]卢辉斌,徐刚,李段.一种基于孤立点检测的入侵检测方法[J].微机发展,2005,15(6):93~94.
[40] Kuramitsu..K../Finding.Periodic.Outliers.over.a.Monogenetic.Event.Stream[C]..P-roc..of.International.Workshop.on.Ubiquitous.Data.Management..Tokyo,.Japan:[S.n.],.2005: 97~104.
[41]杨永铭,王喆.孤立点挖掘算法研究.计算机与数字工程,2008(1):11~15.
[42] Sherhar..S.,.Lu..C.,.Zhang.P...A.unified.approach.to.detecting.spatialoutliers[J]..Geolnfomlatica,.2003.
[43] Breunig. M. M.,.Kriegel. H. P., Ng. R., et al. LOF: Identifying Density-Based Local Outliers[C]. Dallas, Texas: Proc. of ACM SIGMOD Conf, 2000: 201~235.
[44]杨风召,朱扬勇,施伯乐.IncLOF:动态环境下局部异常的增量挖掘算法[J].计算机研究与发展,2004,41(3):477~484.
[45]蒋宗礼.人工神经网络导论.高等教育出版社,2001.
[46] Shavilik. J. W.,.Mooney. R. J...Symbolic.and.neural.learning.algorithms..Maching. Learning,.1991(6): 111~114.
[47] Weiss..S.,.Kapouleas..I...An.empirical.comparison.of.pattern.recognition,.neural. nets.and machine learning classing classification methods. Proceedings of the Eleventh IJCAI, 1989.
[48] Vonder Malshurg. C. Sels-organizing of orientation sensitive cells in the striate cortex. Kybernetikl. 1973.
[49] Lippmann. R. P.. An Introduction to Computing With Neural Nets[M]. IEEE ASSP Magazine, 1999: 23~45.
[50] Cyberko. G.. Approximations by superpositions of a sigmoidal function[A]. Math Control Singnal System[C], 1989.
[51] Hecht-Nielsen. R. Theory of back propagation neural network[J]. Proc of IJCNN, 1989(1): 593~603.
[52] Kennedy..J.,.Eberhart..R..C...Particle.swarm.optimization[A]..Proc..IEEE.Int..Co-nf. on Neural Networks[c], Perth WA Australia, 1995.
[53] Eberhart..R..C.,.Kennedy..J..A..new.optimizer.using.particle/swarm.theory[A]..Pr-oc.the.Sixth.lnt..Symposium.on.Micro.machine.and.Human.Science[C]..Nagoya, Japan, 1995: 39-43.
[54]谢晓锋,张文俊,杨之廉.微粒群算法综述.控制与决策,2003(3):129~134.