A deep-SIQRV epidemic model for COVID-19 to access the impact of prevention and control measures.

The coronavirus (COVID-19) has mutated into several variants, and evidence says that new variants are more transmissible than existing variants. Even with full-scale vaccination efforts, the theoretical threshold for eradicating COVID-19 appears out of reach. This article proposes an artificial intelligence(AI) based intelligent prediction model called Deep-SIQRV(Susceptible-Infected-Quarantined-Recovered-Vaccinated) to simulate the spreading of COVID-19. While many models assume that vaccination provides lifetime protection, we focus on the impact of waning immunity caused by the conversion of vaccinated individuals back to susceptible ones. Unlike existing models, which assume that all coronavirus-infected individuals have the same infection rate, the proposed model considers the various infection rates to analyze transmission laws and trends. Next, we consider the influence of prevention and control strategies, such as media marketing and law enforcement, on the spread of the epidemic. We employed the PAN-LDA model to extract features from COVID-19-related discussions on social media and online news articles. Moreover, the Long Short Term Memory(LSTM) model and Evolution Strategies(ES) are used to optimize transmission rates of infection and other model parameters, respectively. The experimental results on epidemic data from various Indian states demonstrate that persons infected with coronavirus had a more significant infection rate within four to nine days after infection, which corresponds to the actual transmission laws of the epidemic. The experimental results show that the proposed model has good prediction ability and obtains the Mean Absolute Percentage Error(MAPE) of 0.875%, 0.965%, 0.298%, and 0.215% for the next eight days in Maharashtra, Kerala, Karnataka, and Delhi, respectively. Our findings highlight the significance of using vaccination data, COVID-19-related posts, and information generated by the government's tremendous efforts in the prediction calculation process.
Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
(Copyright © 2023 Elsevier Ltd. All rights reserved.)