Аннотация:
This paper is dedicated to investigating the transmission and prediction of viruses within human society. In the first phase, we augment the classical Susceptible-Exposed-Infectious-Recovered (SEIR) model by incorporating four novel states: protected status ($P$), quarantine status ($Q$), self-home status ($H$), and death status ($D$). The numerical solution of this extended model is obtained using the well-established fourth-order Runge-Kutta algorithm. Subsequently, we employ the next matrix method to calculate the basic reproduction number ($R_0$) of the infectious disease model. We substantiate the stability of the basic reproductive number through an analysis grounded in Routh-Hurwitz theory. Lastly, we turn to the application and comparison of statistical models, specifically the Autoregressive Integrated Moving Average (ARIMA) and Bidirectional Long Short-Term Memory (Bi-LSTM) models, for time series prediction.