<p>Recently, mathematical models, which are expressed by a set of ordinary differential equations, which describe the transmission of infectious diseases, have been widely used. The parameters included in this mathematical model are the key elements that characterize the state of transmission of the epidemic. Therefore, it is important to set the parameters realistically. This work aimed to develop a method to determine the unknown parameters of mathematical model of epidemic transmission by genetic algorithm. This work proposed a method for determining unknown parameters in a model by genetic algorithms, one of the global optimization calculations. To test the validity of this method, a SEIHRD model was built and the accuracy of the simulation was compared with the results of determining unknown parameters when the underlying data were prepared in various forms. Because the equations are coupled together, a number of unknown parameters can be determined with a few statistic data. This method can be a recursive method to simply determine unknown parameters in a mathematical model study considering the epidemic transmission situation. The contribution of the paper is that it is possible to approximate the parameters involved in the whole model, even with the observed data of some of the groups in the model containing several groups. In other words, even with incomplete measured data, all parameters of the model can be determined approximately. In the SI model, the absolute error of the parameter does not exceed 0.0104, the relative error does not exceed 1.8484, and in the SEIHRD model using the measured data of S, H, R, the MSE does not exceed 0.0101·10^–5.</p>