ARIMA模型在深圳人口与医疗需求预测的应用

The Application of ARIMA Forecasting Model in Population and Medical Needs of Shenzhen

本文利用SAS/ETS统计软件对深圳1979～2010年常住人口拟合了ARIMA(1,1,0)模型,并由此模型得到未来10年深圳常住人口的预测数据。同时对深圳1979～2010年常住人口每千人床位数拟合了带有ARIMA(1,0,0)误差的线性回归模型。以上的两个模型对1979～2010年的实际数据拟合效果都比较好,基本能反映出两者的发展规律。通过模型预测,未来10年深圳的常住人口数量将超过1300万人,随着人口的自然增长,深圳市的每千人床位数量会继续保持在相对较低的水平上,市民就医压力仍会较大。

This paper establishes the ARIMA（1,1,0） model for Shenzhen resident population in 1979-2010 by using SAS/ETS statistical software,and forecasts the resident population in the next 10 years of Shenzhen based on it.Furthermore,the linear regression model is established with ARIMA（1,0,0） error for the number of beds per 1000 resident population in 1979-2010 of Shenzhen.The two above models fit the real data of 1979-2010 years very well,so they can be used to describe the development law of the resident population and the number of beds per 1000 resident population.By the above forecasting models,it is concluded that the Shenzhen resident population will be over 1.3 million in the next 10 years,and the number of beds per 1000 resident population will be continued in a relative low levels as the population natural growth.The residents will still feel some pressures on medical treatment.