Optimizing COVID-19 Epidemiological Models: A Particle Swarm Approach to Parameter Estimation

Abstract

The SEIR (Susceptible-Exposed-Infectious-Recovered) mathematical model, represented as a system of nonlinear differential equations, has proven to be a powerful tool to describe the dynamics of the spread of infectious diseases such as COVID-19. The accuracy of the projection and understanding of this model relies heavily on the proper estimation of its parameters, such as transmission rate, incubation rate, natural birth rate, natural death rate, disease mortality rate, and recovery rate. This study focuses on the development and application of a new approach to estimate crucial parameters in the SEIR model by utilizing Particle Swarm Optimization (PSO). PSO is a metaheuristic optimization algorithm inspired by the social behavior of flocks of birds or schools of fish. PSO was chosen for its outstanding ability to find a global minimum in a complex search space, as well as its efficiency in handling nonlinear optimization problems. The advantage of PSO lies in its effective memory capacity, which allows the storage of previous best values, both individually and globally, thus accelerating convergence to the optimal solution. Through a simulation program, this study successfully identified the optimal set of parameters for the SEIR model. These estimated parameters were then carefully evaluated by comparing the model simulation outputs with available COVID-19 epidemiological data, demonstrating the model's ability to accurately replicate pandemic trends. The results of this study are expected to make a significant contribution to modelling and understanding the spread of COVID-19.