Comparison of ARIMA and LSTM for Prediction of Hemorrhagic Fever at Different Time Scales in China

Abstract ObjectivesThis study intends to build and compare two kinds of forecasting models at different time scales for hemorrhagic fever incidence in China.MethodsARIMA and LSTM model were adopted to fit monthly, weekly and daily incidence of hemorrhagic fever in China from 2013 to 2018. The two models, combined and uncombined with rolling forecast, were used to predict the incidence in 2019 to identify its stability and availability. ResultsARIMA (2, 1, 1) (0, 1, 1)12, ARIMA (1, 1, 3) (1, 1, 1)52 and ARIMA (5, 0, 1) was selected as the best fitted ARIMA model for monthly, weekly and daily incidence series respectively. The model with 64 neurons and SGDM for monthly incidence, 8 neurons and Adam for weekly incidence, and 64 neurons and RMSprop for daily incidence were selected as the best fitted LSTM models. The values of root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) of the models combined with rolling forecast in 2019 were lower than those of the direct forecast models for both ARIMA and LSTM. It was shown from the forecasting performance in 2019 that ARIMA was better than LSTM for monthly and weekly forecasting while the LSTM was better than ARIMA for daily forecasting in rolling models.ConclusionsBoth ARIMA and LSTM could be used to build a prediction model for the incidence of hemorrhagic fever meanwhile different models might be more suitable for the incidence prediction at different time scales.