A Moving Average Genetic Algorithm (MA-GA) for Estimating the COVID-19 Dynamic Based on a Stochastic SIRD Model

In this study, we examine the transmission of the COVID-19 outbreak using a constructed SIRD stochastic model. To determine the most appropriate model parameters, three stochastic models are proposed, and genetic algorithms (GA) are employed. However, the standard GA has proven inadequate in obtaining suitable parameters for the model, leading to occasional discrepancies in tracking trends from actual case data. To overcome this limitation, we propose a novel modification of the genetic algorithm, termed the Moving Average Genetic Algorithm (MA-GA). Unlike the standard GA, our MA-GA continuously updates the parameters at predetermined intervals, resulting in significantly improved accuracy. By applying this method, we achieve higher precision in providing solutions for the stochastic SIRD model, thereby enhancing its ability to accurately reflect the real-world dynamics of the COVID-19 outbreak.