基于ARIMA模型的重庆市流行性感冒预测研究

Prediction of influenza in Chongqing，China，based on the Autoregressive Integrated Moving Average model

摘要:

目的 分析流感的流行趋势，构建流行性感冒发病的自回归滑动平均混合模型（autoregressive integrated moving average，ARIMA）并对验证集进行预测，为重庆市流感的防治提供科学依据。方法 本研究采用R软件对重庆市2010年1月至2021年6月的流感数据进行ARIMA模型拟合，用2021年7月至2021年12月的数据评价模型拟合效果。结果 该流感疾病的流行呈现出明显的季节周期性，周期为1年，发病高峰期为冬季和春季，整体患病率呈上升后有下降趋势，拟合最佳模型为ARIMA（0，1，2）×（0，1，2）₁₂，且该模型在预测2021年7月至12月罹患率的均方根误差为10.70，平均绝对百分比误差为70.04%，预测效果较好。结论 ARIMA模型在预测重庆市流感发病和流行趋势上有一定的效果，并能对未来的罹患率进行估计，可以为今后的流感疾病防控工作提供参考。

Abstract:

Objective To investigate the trend of influenza prevalence by constructing an Autoregressive Integrated Moving Average （ARIMA） model for influenza and making predictions on the validation set，and to provide a scientific basis for the prevention and control of influenza in Chongqing，China.Methods In this study，R software was used for the ARIMA model fitting of influenza data in Chongqing from January 2010 to June 2021，and the data from July to December 2021 were used to evaluate the fitting performance of the model.Results The prevalence of this influenza disease presented with a noticeable seasonal pattern，with a yearly cycle and a peak in winter and spring. The overall prevalence rate tended to increase first and then decrease，and the best-fitting model was ARIMA（0，1，2）×（0，1，2）₁₂，which had a root mean square error of 10.70 and a mean absolute percentage error of 70.04% in predicting the attack rate in July to December 2021，suggesting that the model had good predictive efficacy.Conclusion The ARIMA model has a certain effect in predicting the onset and prevalence trend of influenza in Chongqing and can estimate the attack rate of influenza in the future，which can provide a reference for the prevention and control of influenza in the future.