Prediction of COVID-19 pandemic based on data-driven model
Download as PDF
DOI: 10.23977/IEMM2021.008
Corresponding Author
Qiyang Sun
ABSTRACT
In order to predict the COVID-19 outbreak, several epidemiological models are used around the world to predict the number of infections and mortality. An accurate predictive model is essential to take appropriate action. According to the latest population migration data in Wuhan, China around January 23, 2020. A data-driven epidemic and prediction method is proposed, and COVID-19 epidemiological data is imported into the Susceptible Exposure Infection Elimination (SEIR) model to derive the epidemic curve. The logistic regression method is used to predict the spread of the virus over time. For further comparison and verification, the LSTM time series model is established to study the trend of virus spread and predict the spread of COVID-19. The epidemic in China should reach its peak in late February and gradually decline by the end of April. The machine learning prediction method adopted in this paper confirms this result and has certain reference significance.
KEYWORDS
COVID-19 prediction, data-driven, machine learning, SEIR, LSTM Networks