时间序列分析法预测高温干旱期间住院病例数量

Time series analysis for prediction of the inpatient-numbers during hot and drought months

目的探讨重庆市局部地区高温干旱期间住院病例数量变化趋势,为本市今后建立与完善高温干旱期间住院病例数量预警系统提供依据。方法通过分析重庆市局部地区高温干旱期间住院病例数量的变化规律,使用时间序列分析法中的ARI-MA模型分析和预测住院病例的变化趋势。结果建立了ARIMA(p,d,q)模型,绝大多数实际值均落在两条预测值的95%置信线内,且该模型进行了短期的预测。结论时间序列分析法用于高温干旱期间住院病例数量变化的研究中具有较好的适应性和适用性,ARIMA(p,d,q)模型能较好的反应并预测重庆市局部地区该期间病例数的变化趋势。

Objective To investigate the variation trend of the number of inpatients during hot weather and drought period,in order to explore how to build and improve early warning system of the amount of inpatients during hot weather and drought months. Methods Through analyzing the variation regularity of time series,ARIMA（p,d,q）model was used to analyze the dynamic variation of the amount of inpatients. Results The modal ARIMA（p, d, q）was established, the vast majority of the actual values fell to the 95% confidence limits of two predictive line,and it produced a good fitness to predict the amount of inpatients in a short future. Conclusion ARIMA（p,d,q）model was applicable and practical in analyzing the variation regularity of the amount of inpatient,which indicate the model ARIMA（p,d,q）can response and predict well the dynamic variation of the amount of inpatients.