Prediction of the COVID-19 Spread in China Based on Long Short-Term Memory Network

Abstract

The sudden outbreak of COVID-19 has caused great losses to the economy and the life of the masses. Long short-term memory (LSTM) network is a time recursive neural network, which is suitable for processing and predicting important events with relatively long interval and delay in time series. Using LSTM network to predict and analyze the development trend of epidemic situation, it is imperative to prevent epidemic situation from causing secondary harm to China’s development. In this paper, we first obtained the COVID-19 data published by China Health Net using crawler technology, which is the accurate value of infection trend after the outbreak of COVID-19 in China. Then, based on these data, the LSTM model is used to predict the development trend of the epidemic in one year, and the mean square error is used to calculate the error between the prediction and the real data. The experimental model is used to predict and analyze the development trend of COVID-19. The results show that the error between predicted data and real data is small and the effect is very good, which provides a reasonable basis and forecast for scientific prevention and control of epidemic situation.