Differential Evolution to Optimize A Long-term Multi-strain Model of COVID-19 in South Africa

A limited number of studies have been conducted to investigate the dynamics of COVID-19 disease spread in South Africa and these existing studies have mostly focussed on mathematical analysis of a relatively short time period near the initial outbreak of COVID-19 in South Africa. The current study therefore attempted to extend on previous studies by applying a Susceptible- Exposed - Infected - Removed (SEIR) disease model to analyse the long-term dynamics of COVID-19 in South Africa, taking into account multiple waves of infection potentially caused by different virus strains. A Differential Evolution (DE) algorithm was used to fit the proposed model to real-world data, and this was done on both a geographically local and global scale to investigate the differences between these two approaches. Results revealed that a local approach provided a more accurate model fit to data than a global approach and that the method proposed in this work could give valuable insights into disease dynamics.