Prediction of Epidemic Situation in COVID-19 Based on Time Series Neural Network

To reveal the transmission rule of COVID-19 in China, the time series neural network prediction model was established to predict the daily cumulative confirmed cases and death cases in the whole nation, Wuhan and Beijing from February 12 to April 15, 2020 respectively, and the prediction performance of the model was evaluated by the relative error between the predicted value and the actual value. The results show that compared with other prediction models, the predicted value of existing cumulative infection cases of the time series neural network prediction model is closer to the actual value, the mean absolute error and the root mean square error are both the smallest, and the prediction accuracy is the highest. The predicted values, the daily cumulative confirmed cases and death cases in the whole nation, Wuhan and Beijing, are more consistent with the actual values with the maximum relative errors of 2.0% and 2.5% respectively, and the prediction model of time series neural network has high accuracy.