Estimation of SIR Epidemiological Model Parameters Using Computational Intelligence Methods Based on COVID-19 Data from Brazil

This study investigates the application of computational intelligence (CI) methods to analyze the transmission dynamics of COVID-19 in Brazil. Utilizing a Susceptible-Infected-Recovered (SIR) epidemiological model based on differential equations, COVID-19 cases are estimated from February 2020 to February 2023. The CI methods employed in this study include Global Random Search, Local Random Search, Harmony Search, Differential Evolution, and Particle Swarm Optimization. The results indicate that Differential Evolution, particularly when applied bimonthly, yielded the most accurate parameter estimations for the model.