Fuzzy Time Series Algorithm for Predicting Epidemic Trends in Infectious Diseases

The spread of epidemics is influenced by a variety of factors, some of which remain unknown. Researchers face uncertainty in building concrete models. This paper introduces the fuzzy time series algorithm, which does not rely on a specific model and starts with data. The algorithm is optimized during the training and testing stages to minimize the mean square error between predicted and actual values. Once the parameters are optimized, they can accurately predict future values. The results of numerical simulations of dengue and COVID-19 demonstrate the superior accuracy of the proposed algorithm compared to other methods. The proposed model has significant advantages in handling fuzzy and complex nonlinear data, effectively predicting the epidemic trends of infectious diseases.