Predicting Enterovirus Epidemics Based on Ensemble Learning

Resumo

The This paper introduced a model aimed at predicting factors that increase incidence of enterovirus infection" case occurrence" among children  less than five years depended to ensemble learning (EL) algorithm.    The research is based on a dataset collected from Al-Khansa Hospital in Mosul, which focuses on paediatric cases of confirmed enterovirus infections and their associated influencing factors. The main goal is to enhanced of decision-making methods to predict this epidemic disease early to support primary care physicians. In this approach, the principal object of this paper is to build model depended to dataset   to enable predict enterovirus epidemics. Following data pre-processing, ensemble learning models are trained to learn patterns that indicate enterovirus incidence, enabling early detection of potential epidemic situations. The model employed AdaBoost, Gradient Boosting, XGBoost, LightGBM, Cat-Boost, stacking and hybrid stacking were used to predict the enterovirus epidemic situation. Experiment results indicate the best-performing model was hybrid stacking achieving an R2 score of approximately 0.77, indicating that ensemble learning models can effectively predict enterovirus epidemics title.