用户名: 密码: 验证码:
基于神经网络的组合预测法在医院管理工作中的应用
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
目的 众所周知,随着中国加入WTO后,东西方的交流日益广泛,西方先进的技术和管理理念必然会对国内市场造成一定冲击,国内原有的医院格局将受到严峻的挑战,医院要在发展中立足,就必须对医院的信息化管理提出更高、更新的要求,必须加快医院管理的现代化、科学化、规范化进程。这也就要求统计工作者能够充分利用已有的时间序列资料,进行正确的统计分析和预测,给管理者提供量化的分析预测结果,以此作为今后工作方向的理论指导,避免管理工作的盲目性。本文旨在通过对医院统计工作中的年度门诊人次和季度入院人次资料进行预测,以寻求一种较为适合医院时间序列资料的预测方法。
     方法 第一部分:收集了天津市某医院的1985年~2002年的门诊人次资料。针对该资料的预测步骤为:首先给出四种不同的单项预测模型:三次指数平滑模型、三次多项式回归模型、灰色模型GM(1,1)、Box—Jenkins模型。
     然后,采用人工神经网络(Artificial Neural Network,简称ANN)将以上这四种模型的拟合预测结果作为网络的输入,以实际的医院门诊人次作为目标输出,建立基于神经网络的组合预测模型。同时以这四种方法的最优加权组合预测模型的结果作为对照,比较变权重与固定权重这两种方法的组合预测效果。
     第二部分:按季度收集了天津市某医院的1996年~2001年的入院人次资料。首先采用季节周期回归模型、温特斯(P.R.Winters)线性和季节指数平滑法、Box—Jenkins模型对该资料进行预测,然后采用人工神经网络将以上三种模型的预测结果进行组合预测。
     第三部分:选取《数理医药学杂志》2000年13卷1期中的“最优加权组合预测法在河北省卫生人力预测中的应用”一文中的卫生人力资料,采用基于人工神经网络的组合预测法进行分析预测,并与该文提出的最优加权组合预测法的预测结果进行比较。
     结果 本研究中对门诊人次资料采用的四种单项预测模型按照预测结果的平均相对误差排序,由小到大依次为三次多项式回归模型
    
    (0*469)、灰色模型GM(.1)(0*655)、Box一Jenkins 模型(0.0743)
    和三次指数平滑模型(0.0789)。
     采用最优加权组合预测模型将这几种模型进行组合预测,结果显
    示预测精度有一定程度的提高(平均相对误差一0刀543)。但是由于这
    几种方法未能通过 kendall一致性检验,K‘-8*25,P>0.05,目单项
    模型之间不具有一致性,因此我们寻求其他的组合预测方法。最后采
    用基于神经网络的组合预测法对门诊人次进行预测,平均相对误差为
    9.996X 105
     另外,在对季节入院人次资料的预测中,基于神经网络的组合预
    测方法也显示了良好的预测结果,平均相对误差为 9.988 X 10”。
     结论 人工神经网络的特点在于它具有高度的自组织学习能
    力,可揭示数据样本中所蕴涵的非线性关系,能以任意精度逼近任意
    函数。而且在ANN模型中,各单项预测模型的权系数不再是一常量,
    而是随时间变化的函数;某一种预测方法对组合预测结果的影响虽然
    与自身的预测结果和权重有关,但其权重对组合预测结果的影响是非
    线性的。这就克服了最优加权组合预测法权重固定不变等缺陷,因此
    基于神经网络的组合预测法实质上是一种变权重系数的组合预测方
    法。变权重系数组合预测模型的建立和应用是提高预测模型的预测精
    度,增强预测模型实用性的有效途径。
     综上所述,在非线性时序资料预测中,基于神经网络的组合预测
    法展现出了它不可替代的强大优势,可以作为统计预测分析的一个得
    力助手在医院管理统计工作中加以应用。
Objective It is known to all, with China being joined the WTO, the exchanges of west-east being increasingly extensive, Western advanced technique and the management principle will inevitably impact domestic local market, and the structure and form of the domestic original hospitals will suffer the rigorous challenge. Then it must put on the pace of scientific hospital management modernization. And this also will require the statistician can make the most of the historical time series data of hospital to analysis and predict, and provide the quantitative forecasting results to hospital manager, which is regarded as the theories guide of the future work direction, avoiding the blindness of management work. The objective of this paper is to explore an appropriate and practical forecasting model for time series data of hospital, through forecasting the yearly data of outpatient numbers and the seasonal data of inpatient numbers.
    Methods Part One: In this study the data of outpatient numbers were collected from 1985 to 2002 in the some hospital in Tianjin City. The forecasting steps of the data were as follows: First, four different single forecasting models were given: cubic exponential smoothing model, cubic polynomial regression model, grey model and Box-Jenkins models. Then, it was set up a nonstationary weights mix forecasting model based on artificial neural network (ANN) with three layers. The inputs of ANN were the forecasting values of above four kinds of methods and output was the actual time series. At the same time, the result of the non-stationary weights mix forecasting model was compared with that of the stationary weights mix forecasting model, in contrast with the optimal weighted mix forecasting model.
    Part Two: In this study the data of inpatient numbers were collected from 1996 to 2001 by four seasons in the some hospital in Tianjin City. First, three different single forecasting methods were given: seasonal regression model, Winters' exponential smoothing method and Box-Jenkins models. Then ANN was applied to mix forecasting according to the results of above three models.
    Part Three: It was applied the health manpower data adopted from the article entitled: "Application of the optimal weighted mix forecasting model in the health manpower in Hebei province" in <> Vol.13, No.l, 2000. And the result of the mix forecasting model based on ANN was
    
    
    
    compared with that of the optimal weighted mix forecasting model.
    Results The order according to the mean relative percent error of forecasting result from low to high was cubic polynomial regression model (0.0469), grey model (0.0655), Box-Jenkins models and cubic exponential smoothing model (0.0789). These single models were applied to build the optimal mix forecasting model. The combination results showed that forecasting precision was improved to some extent, relative error was 0.0543. But these methods could not pass the kendall consistency
    test (x2=8.625, P>0.05), namely they did not have the consistency among these
    single models, so we looked for another mix forecasting method. Finally it was constructed the mix forecasting method based on ANN to forecast the numbers of outpatient, relative error was 9.996 10-5.
    In addition, artificial neural network also seemed to provide adequate forecasts for the seasonal data of inpatient numbers, relative error was 9.988
    10-5.
    Conclusion ANN has high self-study ability and can show the nonlinear relation among the data of sample with strong ability to approximate functions. And the weight coefficient of any single model in ANN is no longer a constant, but is a function with time. Although the influence that a single method work on the result of mix forecasting method is relevant to the forecasting results and weights of itself, it is a nonlinear relation. The mix forecasting method based on ANN overcame the limitation of stationary weights on optimal mix forecasting model, therefore it is a mix forecasting method of nonstationary weights in nature. It
引文
1.邓丹,王润华,周燕荣.时间序列分析及其在卫生事业中的应用.数理医药学杂志,2002,15(5):455-457.
    2.李引珍,盖宇仙.一种多目标模糊组合预测法.铁道学报,1993,15(3):122-125.
    3.王启栋,王洁贞,刘荣甫.现代卫生事业中的统计预测.中国卫生统计,2001,18(4):245-248.
    4.蒋惠园,杨大鸣.货运量预测方法的比较.运筹与管理,2002,11(3):74-79.
    5.程佚.常用预测方法及评价综述.四川师范大学学报(自然科学版),2002,25(1):70-73.
    6.纪爱兵,孙建平,张艳娥.最优加权组合预测法在河北省卫生人力预测中的应用.数理医药学杂志,2000,13(1):61-62.
    7.王景,刘良栋,王作义.组合预测方法的现状和发展.预测,1997,(6):37-38.
    8.陈华友,侯定丕.基于预测有效度的优性组合预测模型研究.中国科学技术大学学报,2002,32(2):172-180.
    9. Yeung LC, James HS, Mark WW, et al. A dynamic factor model framework for forecast combination. Spanish Economic Review, 1999, (1): 91-121.
    10.徐强.统计学中的组合思想初探.浙江统计,2002,(2):20-21.
    11.梁慧稳,王慧敏.经济预测方法系统研究.现代管理科学,2002,(6):28-31.
    12.杨桂元,唐小我.非负权重组合预测模型优化方法研究.数量经济技术经济研究,1998(3):56-60.
    13.蒋良奎.一种在组合预测中确定变权系数的方法.上海海运学院学报,2002,23(3):79-81.
    14.孟建良,王晓华.全局时变权组合预测方法.计算机工程与应用,2002,38(10):98-99.
    15. Franchini L, Spagnolo C, Rossini D, et al. A neural network approach to the outcome definition on first treatment with sertraline in a psychiatric population. Artificial Intelligence in Medicine, 2001, 23(3): 239-248.
    16.卢纹岱主编.SPSS for Windows统计分析.北京:电子工业出版社,2000.
    17.许东,吴铮编著.基于MATLAB 6.x的系统分析与设计—神经网络(第二版).西安:西安电子科技大学出版社,2002,4-24.
    18.章扬熙编著.医学统计预测.北京:中国科学技术出版社.1995,103-108,61-64.
    
    
    19.向前,胡平玲.指数平滑法在医院统计预测中的应用.96全国卫生统计学术研讨会论文集.1996,19-21.
    20.余以双.应用三次指数平滑法预测医院住院人次.中国卫生统计,1996,13(4):33-34.
    21.冯刘栋.试用三次指数平滑法预测传染病发病率.数理医药学杂志.2000,13(2):146.
    22.林昆,陈少如.鲩鱼胆汁染毒大鼠全血还原型谷胱甘肽动态变化及其数学模型.汕头大学医学院学报,1991,(1):44-48.
    23.吴淑艳,王伟,汪培山,等.三次抛物线模型在医院门诊和住院工作量预测中的应用.中国医院统计,1998,5(1):7-10.
    24.余文杰,汤大俊.应用灰色模型预测成都铁路局传染病发病率.预防医学情报杂志,2001,17(4):234.
    25.严金燕,严红艳,杨超.应用灰色模型对门急诊接诊人数进行分析预测.中国医院统计,2001,8(3):152-153.
    26.姚莉.灰色数列预测模型在传染病死亡率研究中的应用.数理医药学杂志,2002,15(2):103-104.
    27.George EPB, Gwilym MJ, Gregory CR. Time Series Analysis Forecasting and Control..顾岚主译.时间序列分析:预测与控制.北京:中国统计出版社,1999,101-148.
    28. Slini T, Karatzas K, Moussiopoulos N. Statistical analysis of environmental data as the basis of forecasting: an air quality application. Science of the Total Environment, 2002, 288(3): 227-237.
    29. Diaz J,Garcia R, Velazquez CF, et al. Effects of extremely hot days on people older than 65 years in Seville (Spain) from 1986 to 1997. International Journal of Biometeorology, 2002, 46(3): 145-149.
    30.倪宗瓒,巫秀美,姚树祥,等.应用ARIMA模型动态分析高危人群的肺癌发病率.数理医药学杂志,2001,14(4):294-296.
    31. Robert SP, Daniel LR. Econometric Models and Economic Forecasts. Mcgraw-Hill, 1998, 538-548.
    32.刘健,黄镇南,孙振球.Box—Jenkins模型在出生人口数短期预测中的应用.中国卫生统计,1992,9(3):44-46.
    33.凌莉,方积乾,汤泽群,等.时间序列方法在卫生人力资源需求预测中的应用.中
    
    国卫生统计,1999,16(5):266-268.
    34.Peter JB, Richard AD. Time Series: Theory and Methods(Second Edition). Springer—Verlag.田铮译.时间序列的理论与方法.北京:高等教育出版社,2001,214-256.
    35.张晓峒著.计量经济分析.北京:经济科学出版社,2000,88-102.
    36.王启栋,刘荣甫,王洁贞,等.用最优加权组合预测法预测济南市人口.中国卫生统计,2001,18(3):179-180.
    37.徐俊彦,苗壮.我吉林省粮食产量最优组合预测.吉林工学院学报,1998,19(3):77-80.
    38.陈欣,张志尧,毕光忠,等.天津市卫生人力资源预测.天津医科大学学报,2000,6(3):284-288.
    39.耿奎.最优加权组合方法探讨.数量经济技术经济研究,1998,(11):44-46.
    40.吴拥军,吴逸明,张振中,等.基于人工神经网络的“最优标志物群”在肺癌诊断中的应用研究.实用肿瘤杂志,2002,17(5):317-320.
    41. Dorsey SG, Waltz CF, Brosch LM, et al. A Neural Network Model for Predicting Pancreas Transplant Graft Outcome. Diabetes Care, 1997, 20(7): 1128-1133.
    42. Somers, Mark J. Application of Two Neural Network Paradigms to the Study of Voluntary Employee Turnover. Journal of Applied Psychology, 1999, 84(2): 177-185.
    43. Sch6nbach C, Kun Y, Brusic V. Large-scale computational identification of HIV T-cell epitopes. Immunology and Cell Biology, 2002, 80(3): 300-306.
    44. Baskin Ⅱ, Ait AO, Halberstam NM, et al. An approach to the interpretation of backpropagation neural network models in QSAR studies. Sar & Qsar in Environmental Research, 2002, 13(1): 35-41.
    45. Mecocci P, Grossi E, Buscema M, et al. Use of Artificial Networks in Clinical Trials: A Pilot Study to Predict Responsiveness to Donepezil in Alzheimer's Disease. the American Geriatrics Society, 2002, 50(11): 1857-1860.
    46.蔡航.基于神经网络的医疗诊断专家系统.数理医药学杂志,2002,15(4):291-295.
    47.贺宪民,贺佳,范思昌.BP神经网络及其预测性能探索.数理医药学杂志,2001,14(3):195-198.
    48.丁守銮,王洁贞,胡平.基于动态学习比率BP神经网络的时间序列预测方法.中国卫生统计,2002,19(4):194-198.
    49.方积乾,陆盈主编.现代医学统计学.北京:人民卫生出版社,2002,711-712.
    50. Jacques DV, Etinne Barnard. Back propagation Neural Nets with One and Two Hidden Layers.
    
    IEEE Trans Neural Networks, 1993, 4(1): 163-141.
    51. Liu HC. Decision Boundary formation form the Back propagation Algorithm. IEEE International Conference on Computer Architecture and DSP, 1989:17-23.
    52.黄德生,刘延令,金一和.BP人工神经网络用于芳香族化合物结构参数和大鼠LD_(50)构效关系研究.数理医药学杂志,2001,14(1):1-6.
    53. Shi shanming, xu Ii D, Liubao. Application of artificial neural networks to the nonlinear combination of forecasts. Expert Systems, August 1996,13(3): 195~201.
    54. Hassibi B, Stork DG, Wolff GJ. Optimal brain surgeon and general network pruning. Proceedings of the IEEE International Joint Conference on Neural Network, 1992, 2: 441-444.
    55.闻新,周露,王丹力,等编著.MATLAB神经网络应用设计.北京:科学出版社,2000,225-232.
    56.陈文娟,郭瑾.应用季节周期回归模型预测甲型肝炎发病率.中国卫生统计,2001,18(6):343.
    57.黄良文主编.统计学原理.北京:中国统计出版社,2000,344-354.
    58.朱士俊,董军.医院管理与信息利用.北京:人民军医出版社.2002,38-40.
    59.夏安邦,王硕编著.定量预测引论.南京:东南大学出版社,2001,43-51.
    60. John EH, Arthur GR, Dean WW. Business Forecasting (Seventh Edition). Prentice Hall, 2001, 74-77.
    61.庞主编.计量经济学.成都:西南财经大学出版社,2001,8-9.
    62.唐芸,秦秀华.一种简易预测模型的应用.林业调查规划,2002,27(2):11-15.
    63.杨学行.三次抛物线配合方法在碘缺乏病区居民患病率分析中的应用.数理医药学杂志,1993,6(2):56-58.
    64.金水高.关于GM(1,1)模型在医学中的误用.中国卫生统计,1992,9(6):42.
    65.许汝福,王文昌,尹金焕,等.时间序列GM(1,1)残差季节周期模型及其应用.数理医药学杂志,1996,9(4):311-312.
    66. Tang W, Zhou S, Yang L. The relationship between chromatographic retention value and grey model parameters. Chinese Journal of Chromatography, 1998, 16(2): 95-99.
    67.刘思峰,郭天榜,党耀国,等著.灰色系统理论及其应用.第二版.北京:科学出版社,2000,102-133.
    68. Tobias A, Diaz J, Saez M, et al. Use of poisson regression and box-jenkins models to evaluate the short-term effects of environmental noise levels on daily emergency admissions in Madrid,
    
    Spain. European Journal of Epidemiology, 2001, 17(8): 765-771.
    69.刘晓宏,陈启明.ARIMA模型中时间序列平稳性的统计检验方法及应用.中国卫生统计,1998,15(3):12-14.
    70. Diaz J, Alberdi JC, Pajares MS, et al. A model for forecasting emergency hospital admissions: effect of environmental variables. Journal of Environmental Health, 2001, 64(3): 9-15.
    71. Schubert C, Lampe A, Rumpold G, et al. Daily Psychosocial Stressors Interfere With the Dynamics of Urine Neopterin in a Patient With Systemic Lupus Erythematosus: An Integrative Single-Case Study. Psychosomatic Medcine, 1999, 61(6): 876-888.
    72. Macciotta NP, Vicario D, Pulina G, et al. Test day and lactation yield predictions in Italian simmental cows by ARMA methods. Journal of Dairy Science, 2002, 85(11): 3107-3114.
    73.徐国祥主编.统计预测和决策.上海:上海财经大学出版社,2001,131-133.
    74.李宝仁.综合预测方法及其改进.北京商学院学报,1996,(6):43-45.
    75.王郁.组合预测何以兴起.预测,1989,(4):22-23.
    76.帅平.基于人工神经网络的GPS单点定位.测控技术,2002,21(10):52-55.
    77. Serra JR, Jurs PC, Kaiser KL. Linear regression and computational neural network prediction of tetrahymena acute toxicity for aromatic compounds from molecular structure. Chemical Research in Toxicology, 2001, 14(11): 1535-1545.
    78. El-Din AG, Simith Dw. A neural network model to predict the wastewater inflow incorporating rainfall events. Water Research, 2002, 36(5): 1115-1126.
    79.张恒喜,郭基联,朱家元,等著.小样本多元数据分析方法及应用.西安:西北工业大学出版社,153-161.
    80. Casadio R, Compiani M, Facchiano A, et al. Protein structure prediction and biomolecular recognition: from protein sequence to peptidomimetic design with the human beta3 integrin. Sar & Qsar in Environmental Research, 2002, 13(3-4): 473-486.
    81. Han M, Snow PB, Brandt JM. Evaluation of artificial neural networks for the prediction of pathologic stage in prostate carcinoma. Cancer, 2001, 91(8 Suppl): 1661-1666.
    82. Drago GP, Setti E, Licitra L. Forecasting the performance status of head and neck cancer patient treatment by an interval arithmetic pruned perceptron. IEEE Transactions on Biomedical Engineering, 2002, 49(8): 782-782.
    83. Park J, Edington DW. A sequential neural network model for diabetes prediction. Artificial Intelligence in Medicine, 2001, 23(3): 277-93.
    84. Walter EP, Steven MW, Ethan R. Use of an artificial neural network to predict length of stay in
    
    acute pallcreatitis. Am Smg, 1998, 64(9): 868-872.
    85. Brickly MR, Shepherd JP, Armstrong RA. Neural networks: a new technique for development of decision support systems in dentistry. J Dent, 1998, 26(4): 305-309.
    86. Davis GE, Lowell WE. Using artificial neural networks and the Gutenberg-Richter power law to "rightsize" a behavioral health care system. Am J Med Qual, 1999,14(5): 216-28.
    87.贺佳,张智坚,贺宪民.肝癌术后无瘤生存期的人工神经网络预测.数理统计与管理,2002,21(4):14-17.
    88.段琼红,聂绍发,仇成轩,等.应用BP神经网络预测前列腺癌流行趋势.中国公共卫生,2000,16(3):193-195.
    89.邓伟,金丕焕.人工神经网络及其在预防医学中的应用.中国公共卫生,2002,18(10):1265-1267.
    90. Bibi H, Nutman A, Shoseyov D, et al. Prediction of Emergency Department Visits for Respiratory Symptoms Using an Artificial Neural Network. Clinical Investigations: Emergency Room, 2002, 122(5): 1627-1632.
    91.谢开贵,李春燕,周家启.基于神经网络的负荷组合预测模型研究.中国电机工程学报,2002,22(7):85-89.
    92.金聪.基于遗传神经网络的癌症死亡率预测.系统工程理论与实践,2000,(2):141-144.
    93.江学军,唐焕文.前馈神经网络泛化性能力的系统分析.系统工程理论与实践,2000,8:36-40.
    94.贺昌政,李晓峰,俞海.BP人工神经网络模型的新改进及其应用.数学的实践与认识,2002,32(4):554-561.
    95.何文章,刘海林.基于神经网络的异位妊娠发病率发展趋势研究.生物数学学报,1999,14(1):72-76.
    96.雷鸣,吴雅,杨叔子.非线性时间序列建模与预测的神经网络法.华中理工大学学报,1993,21(1):47-52.
    97.张青.基于神经网络最优组合预测方法的应用研究.系统工程理论与实践,2001,(9):90-93.

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700