ARIMA模型及其在医疗资源消费预报中的应用

发布时间：2018-05-06 14:53

本文选题：ARIMA模型 + 预报　；参考：《山西医科大学》2008年硕士论文

【摘要】： 时间序列是按照时间顺序取得的一系列观察值,时间序列数据的本质特征就是相邻观察值之间的依赖性。所以在对时间序列数据进行分析时,一般的回归模型难以体现变量自身前后及应变量与自变量过去的依赖关系。自回归求和滑动平均(ARIMA)模型是对变量自身前后依赖性进行分析的技术,它描述了变量自身当前与过去的统计依赖关系,在显示变量的动态系统(dynamical system)演变规律方面有着较为丰富的结构。 ARIMA模型由George E.P.Box和Gwilym M.Jenkins首次系统提出,其建模的假定较少,容易得到满足,在现实的系统中有着广泛的应用。使用ARIMA模型对时间序列进行分析,最大的意义在于预报。本文中,我们将利用ARIMA模型对医疗资源消费进行预报,并将预报结果与实际值和普通回归模型的预报结果进行比较。
[Abstract]:The time series is a series of observation values obtained according to the time sequence. The essential feature of the time series data is the dependence between the adjacent observation values. Therefore, in the analysis of time series data, the general regression model is difficult to reflect the variables before and after and the dependent relationship between dependent variables and independent variables in the past. The autoregressive summation moving average (ARIMA) model is a technique for analyzing the dependence of variables before and after, which describes the current and past statistical dependencies of variables themselves. There are abundant structures to show the evolution of dynamic system. ARIMA model is proposed by George E.P.Box and Gwilym M.Jenkins for the first time. Its modeling assumptions are few and easy to be satisfied. It is widely used in real systems. The ARIMA model is used to analyze the time series, and the greatest significance lies in the prediction. In this paper, we will use ARIMA model to forecast the consumption of medical resources, and compare the forecast results with the actual values and the results of the general regression model.
【学位授予单位】：山西医科大学
【学位级别】：硕士
【学位授予年份】：2008
【分类号】：R311;R197.1

【参考文献】