A susceptible-exposed-infected-quarantined-recovered (SEIQR) model for predicting the trajectory of the COVID-19 epidemic

Li, B. X.; Liu, Z. H.; Zhao, C. H.; Sun, Y. X.; Ieee,; Li, B.; Liu, Z.; Zhao, C.; Sun, Y.

In recent days, the outbreak of COVID-19 poses a great threat to global public health. While great efforts have been made to combat the virus by medical workers of various countries, considerable attention has also been paid to forecast the trend of the epidemic by the researchers, which may help the government formulate corresponding policies. In this paper, we proposed an improved SEIR epidemiological model, named SEIQR (susceptible-exposed-infected-quarantined-recovered). Quarantine group and several time-varying parameters were considered to make the model better approximate the real situation. Besides, ABC-SMC algorithm was introduced to tackle parameter inference problem and ABC-SMC model selection framework was employed to demonstrate the superiority of SEIQR over traditional methods. Extensive experiments have shown that the proposed SEIQR model can perform well in identifying epidemic trends and predicting future development.