“…In temporal order, we divide each patch of these two datasets into training, validation, and test sets at ratios of 60%, 20%, and 20%, respectively, and normalize all data to the range (0, 1). As pointed out in references [50,51], epidemic forecasting often overlooks undocumented cases, and the quality of estimated data impacts subsequent forecasting. This study primarily focuses on analyzing model forecasting accuracy assuming that these data are ideal.…”