基于指数平滑模型的克拉玛依市流感样病例预测分析

Predictive analysis of influenza-like illness based on exponential smoothing method in Karamay city

目的应用指数平滑法预测新疆克拉玛依市流感样病例就诊比(ILI%)变化趋势,为当地流感防控工作提供技术支持。方法利用2018年1 w—2020年52 w的ILI%数据拟合建立指数平滑模型,选择最优模型预测2021年1-26 w ILI%,并以实际值和预测值的相对误差验证模型预测效果。结果 2018—2020年克拉玛依市ILI%呈季节性周期变化;简单季节模型为最优模型,平均相对误差为43.63%,2021年1-9 w平均相对误差为21.83%,第八周相对误差最小为3.62%,10-26 w平均相对误差为55.18%。结论指数平滑模型对ILI%的短期预测效果较好。

Objective To fit and predict the incidence trend of influenza-like illness ratio(ILI%) with exponential smoothing method in Karamay city of Xinjiang, and provide technical support for local ILI prevention and control. Methods The fitting of weekly ILI% data from week 1, 2018 to week 52, 2020 were taked for establishing exponential smoothing models. The optimal model was selected to predict the ILI% of from week 1 to 26, 2021, and the effect of prediction model was verified by the relative errors between the actual values and the predictive values. Results The incidence rates of influenza-like illness showed seasonal variation, and the simple seasonal model was optimal with the average relative error of 43.63% from 2018 to 2020. The mean of relative errors was 21.83% from week 1 to 9, 55.18% from week 10 to 26, and the relative error was minimum(3.62%) in week 8, 2021. Conclusions The effect of exponential smoothing model is good on short-term prediction of ILI epidemic.