Application of ARIMA model based on Python language to predict the incidence of foodborne diseases in Jiangxi Province

Abstract

ObjectiveTo evaluate the feasibility of the autoregressive moving average model （ARIMA） for predicting the monthly incidence of foodborne diseases in Jiangxi Province.MethodsThe ARIMA model was constructed by Python software， and the data from January to December in 2021 was used as the validation set to evaluate the prediction performance of the ARIMA model. The short-term prediction of the monthly incidence of foodborne diseases in Jiangxi Province from January to June in 2022 was carried out.ResultsThe incidence of foodborne diseases in Jiangxi Province from 2016 to 2021 generally showed a downward trend， with the peak incidence in August each year. The best prediction model was ARIMA（1，0，0）（1，0，2）12. The Bess Information Criterion （BIC） was 96.66， and the model residual was a white noise sequence （P>0.05）. The predicted incidence rate of the model was roughly consistent with the actual incidence trend， and the overall root mean square error （RMSE） was 0.656. The efficacy of the model was verified by the data in 2021. The mean absolute percentage error （MAPE） between the predicted value and the actual value was 11.25%. It showed that the model extrapolation effect was better.ConclusionThe ARIMA（1，0，0）（1，0，2）12 model can be used for short-term prediction of the incidence trend of foodborne diseases in Jiangxi Province.

Keywords

• foodborne disease
• arima model
• python language
• prediction