用户名: 密码: 验证码:
最优组合预测法在天津市卫生人力资源需要量中的应用
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 卫生人力资源是反映一个国家或地区卫生服务水平的重要指标之一,是卫生事业中最具有活力的、最积极的因素,开发研究卫生人力资源对提高卫生服务质量具有重要的意义。近些年来,区域性卫生人力资源的研究已经成为世界各国共同关心的课题,我国卫生改革的重要措施之一就是推进区域性卫生人力资源规划。目前,我国卫生人力建设取得了显著的成绩,但也仍然存在着一些问题,主要表现在卫生人力资源配置不合理、卫生机构中人员相对饱和与人才不足共存的现象。因此,这就要求卫生工作者对卫生机构人员的需要量进行预测,为卫生人力的规划和人员配置提供科学的依据。统计预测作为一门新兴的综合科学,经过几十年的发展日趋成熟,已经被广泛地应用于卫生人力资源的规划中。本文旨在通过对天津市卫生技术人员需要量的预测,寻求较为合适的预测方法,并为今后卫生技术人员配置提供依据。
     方法 第一部分:收集天津市1978~2003年的人口数,利用此资料对天津市2004年的人口数进行预测。预测的步骤为:首先采用二次多项式曲线、Box-Jenkins模型和灰色模型GM(1,1)分别对天津市的人口进行预测,并对预测的效果进行评价。然后采用最优加权组合预测方法将上述三种方法进行组合,并将预测的效果与上述三种方法进行比较。最优加权组合预测法是对三种预测方法分别赋予不同的权重,然后将这三种方法进行组合,使预测误差达到最小,三种预测方法的权重合计为1。权重的确定是采用线性规划的方法,使目标函数极小化。
     第二部分:应用天津市1978~2003年的卫生人力人口比值,对2004年的卫生人力人口比值进行预测。预测的步骤为:首先采用Box-Jenkins模型、回归预测法和Brown三次指数平滑法分别对天津市的人力人口比值进行预测,并对预测的效果进行评价。然后采用最优加权组合预测方法将上述三种方法进行组合,并将预测的效果与上述三种方法进行比较。
     第三部分:采用人力人口比值法对2004天津市卫生技术人员的需
    
    硕士研究生学位论文
    要量进行预测。人员需要量=预测年的人口数x卫生人力人口比值数[3]。
     结果在对人口的预测中二次多项式曲线、Box一Jenkins模型和灰色模型
    GM(1,l)的平均相对误差依次为0.37%、0.31%和1.38%,最优加权组合预测
    法的平均相对误差为0.28%,预测精度高于单一预测方法。在对人力人口比值的
    预测中Box一Jenklns模型、回归预测法和Brown三次指数平滑法的平均相对误差
    依次为2.00%、2.79%、1.83%,组合预测的平均相对误差为0.98%,预测精度高于
    单一预测方法。对天津市2004年卫生技术人员需要量的预测结果为:56985人。
     结论目前,预测的方法很多,不同的预测方法能够提供不同的有用信息,
    如果简单地将预测误差较大的一些方法舍去,将会失去一些有用的信息。一种
    较科学的方法就是进行组合预测。组合预测能够较大限度地利用各种预测样本
    信息,比单个预测模型考虑问题更系统全面,能够有效的减少单个预测模型中
    一些随机因素的影响,从而提高预测精度。组合预测法计算简单,精度高,有
    很好的实用性,因此应用其预测卫生人力资源的需求,可以为卫生人力规划提
    供有力的科学依据。
Objective Health manpower resource is one of important indexes reflecting health service standard of a country or a area, as the most active factor, developing and researching it have an significant meaning to advance the quality of health service. In recent years, research about territorial health human recourse has been concerned by all over the world, and advancing it's programming has become a measure of our health innovation. Now, some remarkable achievements have been gained in our health human building, but there are also some questions about it. These mainly reflect in illogical collocation of health manpower that comparatively saturated and the short are coinstantaneous in the health institution. Therefore, it requires health researchers forecast the needs of health institution, which can provide scientific bases for health manpower program and arrangement. As a new compositive science, statistical forecasting has developed consummate through many years, and it has been applied extensively in the la
    yout of health manpower. The objective of this paper is to explore an appropriate forecasting model through forecasting needs of Tianjin health technician, and provide bases for the arrangement in the future.
    Methods Part One: Collect the number of Tianjin population from 1978 to 2003,and use it to forecast population of 2004.The forecasting steps is as follow: First, apply second-degree polynomial curve model, Box-Jenkins model and grey model to forecast population separately and evaluate the forecasting results. Then, combine them by optimum forecast combining, and comparing forecasting results with those of them. Optimal weights for combined forecasts is a model that gives different weight to three single models, then combines them to make error lowest. The weights of three models are add up to one. The weights are determined by linear programming method which can make objective function minimal.
    Part Two: Collect Tianjin health manpower: population ratio from 1978 to 2003, and use it to forecast the ratio of 2004. The forecasting steps is as follow: First, apply Box-Jenkins model, regression forecast model and Brown cubic exponential
    
    
    smoothing model to forecast ratio separately and evaluate the forecasting results. Then, combine them by optimum forecast combining, and comparing forecasting results with those of them.
    Part Three: Apply health manpower: population ratio method to forecast needs of Tianjin health technician of 2004. Needs equal the number of population in forecasting year by manpower: population ratio.
    Results In the forecasting of population, the mean relative percent error of second-degree polynomial curve model, Box-Jenkins model and grey model are 0.37%, 0.31% and 1.38%, and that of optimum forecast combining is 0.28% which forecasting precision is higher than single model. In the forecasting of manpower: population ratio, the mean relative percent error of Box-Jenkins model, regression forecast model and Brown cubic exponential smoothing model are 2.00%, 2.79% and 1.83%, and that of optimum forecast combining is 0.98% which forecasting precision is higher than single model. The total number of health technical staffs of Tianjin of 2004 are expected to be 56985 respectively.
    Conclusion Now, there are many forecasting methods, different method can provide different useful information. If simply give up some methods which have big errors, it may be lost some useful information. Forecast combining is a scientific method. It can largely use information of various forecast methods, which consider plentifully than single forecasting model and reduce effect induced by random factors, then improve forecasting precision. Forecast combining is simple, high precision and good practicality, therefore applying it to forecast the needs of health manpower can provide powerfully scientific bases for health manpower programming.
引文
1.龚幼龙.卫生人力需要量预测方法.中国社会医学,1992,(5):11-12,21
    2.魏颖主编.卫生经济学与卫生管理.北京:人民卫生出版社,1998,147-149
    3.高歌,龚幼龙.上海市2010年卫生人力需要量预测研究.中国农村卫生事业管理,1996,16(9):42-44
    4.唐小我主编.预测理论及其应用研究.西安:电子科技大学出版社,1992,1-68
    5.徐大江.组合预测及其应用研究.统计与预测,1996,(1):22-24
    6.杨广喜.经济预测中组合预测法的应用.统计与预测,1998,(6):17-20
    7.严云良.灰色GM(1,1)模型在卫生资源预测中的应用.数理医药学杂志,1999,12(2):160-161
    8.徐国祥,胡清友编著.统计预测和决策.上海:上海财经大学出版社,2001,133-134
    9.王郁.组合预测何以兴起.预测,1989,(4):1-2,13
    10.李宝慧,项静恬.浅谈权重综合方法及组合预测.统计教育,2000(6):28-29
    11.张世英,张文泉.技术经济预测与决策.天津:天津大学出版社,1994,75-79
    12.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
    13.刘海波,项静恬,陈国珍等.时间参数加权的递归正权组合预测方法及其在气候预测中的应用.数理统计与管理,1999,18(3):35-39
    14.葛雄灿,胡秉民,高毅.最优组合预测方法及其在Logistic曲线与Gompertz曲线之综合拟合中的应用.浙江农业大学学报,1998,24(4):443-446
    15.侯建中,张福林.用最优加权组合法预测深圳市人口发展趋势.数理医药学杂志,1998,11(3):203-205
    16.卢纹岱主编.SPSS for Windows统计分析.北京:电子工业出版社,2000
    17.倪宗瓒主编.卫生统计学第四版.北京:人民卫生出版社,2001,132-136
    18.江之源.经济预测方法与模型.成都:西南财经大学出版社,1999,74-98
    19.张晓峒主编.计量经济分析.北京:经济科学出版社.2000,60-61
    20.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
    21.易丹辉主编.统计预测方法与应用.北京:中国人民大学出版社1990,102-128
    22.顾岚.时间序列分析在经济分析中的应用.北京:中国统计出版社,1994,113-114
    23.张晋昕,何大卫,王亚拉等.ARIMA模型在医院卫生消耗材料需求量预测中的应用.中国医院统计,1999,6(4):210-212
    24.蔡辉,陆守增,倪宗瓒.综合效益值的灰色动态预测.中国卫生统计.2000,17(5):282-284
    25.谢开贵,周家启.变权组合预测模型研究.系统工程理论与实践,2000(7):36—40,117
    
    
    26.项静恬,史久恩编著.非线性系统中数据处理的统计方法.北京:科学出版社,2000,90-128
    27.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
    28.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
    29.朱建中,邵建利.统计应用软件EXCEL和SAS.上海:财经大学出版社,2002,102-115
    30.温庆,杨哲.广东省卫生人力资源现状及发展.统计与预测,1999,(6):39-41
    31.刘国旗,江启成.从我国医学教育规模看卫生人力发展趋势.中国农村卫生事业管理,2000,20(2):31-33
    32.罗文丽,郭岩.卫生人力资源配置研究.《国外医学》医院管理分册,2000,(1):5-7
    33. Arnold E, Pulich M. Managing effectively in the downsized organization. Health Care Manager (Frederick) 2003, 22 (1): 56-62
    34. Dobson R. WHO to create international human resources database on health care. BMJ, 2003 May 10;326(7397): 1004
    35. Persaud DD, Cockerill R, Pink G. et al. Determining Ontario's supply and requirements for ophthalmologists in 2000 and 2005:1. Methods.Can J Ophthalmol, 1999,34(2): 74-81
    36.梁智,常江,徐建国等.卫生人力需求预测方法学比较研究.卫生软科学,1997(2):22-25
    37.杨曾武.统计预测原理.北京:中国财政经济出版社,1990,10-13
    38.夏安邦,王硕编著.定量预测引论.南京:东南大学出版社,2001,49-51
    39. Wenzel Thomas. Hits-and-Misses for the Evaluation and Combination of Forecasts. Journal of Applied Statistics, 2001, 28(6): 759-773
    40.邢志强.指数平滑法预测信息业实例分析.统计与预测,2000,(1):33-35,6
    41. D'Sa MM, Nakagawa RS, Hill DS,et al. Exponential smoothing method for forecast drug expenditures. American journal of hospital pharmacy, 1994,51(20): 2581-2588
    42. Xue JL, Ma JZ, Louis TA, et al. Forecast of the number of patients with end-stage renal disease in the United States to the year 2010. J Am Soc Nephrol, 2001,12(12): 2753-2761
    43. Dewa CS, Hoch JS, Goering P. Using forecasting models to estimate the effects of changes in the composition of claims for selective serotonin reuptake inhibitors on expenditures. Clin Ther, 2001,23(2): 292-306
    44.程昳.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
    45.张文彤.SPSS 11统计分析教程.北京希望电子出版社,2002,271-272
    46. Gross,Kevin D. Forecasting the Future. Journal of Property Management, 2001,
    
    66, (3): 52-56
    47. Williams Dan W, Miller Don. Level-Adjusted Exponential Smoothing for Modeling Planned Discontinuities. International Journal of Forecasting, 1999, 15(3): 273-289
    48. Wilson Patrick J, Okunev John. Comparing Univariate Forecasting Techniques in Property Markets. Journal of Real Estate Portfolio Management, 2000, 6 (3): 283-307
    49. Cirincione Carmen, Gurrieri Gustavo A, Van De Sande Bart. Municipal Government Revenue Forecasting: Issues of Method and Data. Public Budgeting & Finance, 1999, 19 (1): 26-47
    50.胡学锋.B-J法在储蓄预测中的应用研究.数理统计与管理,2001,20(3):49-53
    51. Gill Samantha J, Biging Gregory S. Autoregressive moving average models of crown profiles for two California hardwood species. Ecological Modelling, 2002, 152 (2): 213-227
    52. Box GEP, Jenkins GM, Reinsel GC. Time Series Analysis: Forecasting and Control, 1994
    53. Allard R. Use of time-series analysis in infectious disease surveillance. Bull World Health Organ,1998, 76(4): 327-333
    54. Eyli Elif Derya, Ger Inan. Determination of Stenosis and Occlusion in Arteries With the Application of FFT, AR, and ARMA Methods. Journal of Medical Systems, 2003, 27 (2): 105-121
    55.叶瑰昀,罗耀华,刘勇等.基于ARMA模型的电力系统负荷预测方法研究.信息技术,2002,(6):74-76
    56.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
    57. Abdel-Aal RE, Mangoud AM. Modeling and forecasting monthly patient volume at a primary health care clinic using univariate time-series analysis. Comput Methods Programs Biomed, 1998, 56(3): 235-247
    58. Reis BY, Mandl KD. Time series modeling for syndromic surveillance. BMC Med Inform Decis Mak, 2003, 3(1): 2
    59. Piwowar J M, Ledrew E F. ARMA time series modelling of remote sensing imagery: a new approach for climate change studies. International Journal of Remote Sensing, 2002, 23(24): 5225-5250
    60. Molenaar, Peter C. M. On the Nature of SEM Estimates of ARMA Parameters. Structural Equation Modeling, 2002, 9(3): 347-369
    61. Tong S, Hu W. Climate variation and incidence of Ross river virus in Cairns, Australia: Time-series analysis. Environ Health Perspect, 2001,109 (12): 1271-1274
    62.杨位钦,顾岚编著.时间序列分析与动态数据建模.北京:工业学院出版社.1994,146-149
    
    
    63.李仲来.中国1901-2000年人间鼠疫动态规律.中国地方病学杂志,2002,21(4):292-294
    64.张蔚,张彦琦,杨旭.时间序列资料ARIMA季节乘积模型及其应用.第三军医大学学报,2002,24(8):955-957
    65.杨桂元,唐小我,马永开.关于非负权重组合预测的若干问题的探讨.电子科技大学学报,1996,25(2):210-215
    66.万星火,邢桂荣.组合预测技术在抗生素管理中的应用.数理医药学杂志,1996,9(3):280-282
    67.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
    68.王培麟.用Excel求解线性规划及线性方程组的方法.西北民族学院学报(自然科学版),2002,23(2):37-39,49
    69.晏林.线性规划及其程序设计.文山师范高等专科学校学报,2002,14(1):61-64
    70.唐润霞,林芳立.浅议现代医院统计预测方法的选择.数理医药学杂志,2001,14(1):51-52
    71.王硕,唐小我.组合预测系统综合研究.科学管理研究,2000,18(4):42-45
    72.李海莲,何琳.医院需求的统计预测方法浅探.山西科技,2001,(6):38-40
    73.刘海波,项静恬,陈国珍等.多种正权综合方法在气候预测中的应用比较和改进.数理统计与管理,2000,20(5):34-39
    74. Fischer Ilan, Harvey Nigel. Combining Forecasts: What Information Do Judges Need to Outperform the Simple Average. International Journal of Forecasting, 1999, 15(3): 227-246
    75. Ridley Dennis. Optimal antithetic weights for lognormal time series forecasting. Computers & Operations Research, 1999, 26(3): 189-210
    76.史永,王鹏,张粒子.一种实用的短期负荷组合预测方法.现代电力,2000,17(1):35-40
    77.雷功炎主编.数学模型讲义.北京:北京大学出版社,2001,1-27
    78.唐纪,王景.组合预测方法评述.预测,1999(2):42-43
    79.吴春霞,何勇,蔡建平.组合预测方法及其在粮食产量预测中的应用.农业系统科学与综合研究,2002,18(1):17-19
    80.徐大江.确定最优组合预测权重数的线性规划方法.预测,1993(2):56-57,59
    81.天津市统计局.天津统计年鉴.北京:中国统计出版社.2001~2003
    82.龚幼龙,冯学山.卫生服务研究.上海:复旦大学出版社,2002,205-219
    83.郭建花,唐彪,王京一等.加权平均组合预测模型在河北省卫生人力资源预测中的应用.中国卫生人才,2002,(12):36-37
    84.沈清,李鲁,刘谷琮.浙江省护士规模的预测研究.卫生软科学,2001,15(4):51~53
    85.李秀央,李振洪,蔡雪霞.用EXCEL实现灰色数列模型GM(1,1)的预测.数理医药学杂志,2000,13(4):296-297
    
    
    86.钟晓妮,周燕荣,易静.重庆市卫生人力资源研究.中国卫生经济,2003,22(2):30-32
    87.毛宗福,王永隶,刘继强.我国卫生人力资源及其研究现状.中华医院管理杂志,2003,19(1):12-16
    88. Chant AD.A confusion of roles: manpower in the National Health Service. J. R. Soc. Med, 1998, 91: 63-65
    89. Blatterman J. The future: "Health-care deliver and design: 2000 to 2004. " Archit Rec. 1997, 185 (5): 304
    90. Persaud DD, Cockerill R, Pink G, et al. Determining Ontario's supply and requirements for ophthalmologists in 2000 and 2005:2. A comparison of projected supply and requirements. Can J Ophthalmol. 1999, 34(2): 82-87
    91. Crabtree BF, Ray SC, Schmidt PM, et al. The individual over time: time series applications in health care research. J Clin Epidemiol. 1990, 43(3): 241
    92. Thomas L.Hall. Human Resource for Health: Models for Projecting Workforce Supply and Requirements. First Edition. Geneva: WHO,1994

© 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