Computational intelligence-based model for diarrhea prediction using Demographic and Health Survey data

Diarrhea is one of the leading public health problems and the third main cause of death among young children in developing countries. Solutions to tackling the infectious disease require both preventive and control efforts. However, efforts toward improving the control measures require comprehending the factors associated with diarrhea incidence and the ability to accurately forecast the incidence of the disease. Therefore, the present study develops a diarrhea incidence prediction model based on the 2013 Nigeria Demographic and Health Survey data using artificial neural network. The empirical results of the model indicate that, by using only 44 demographic, socioeconomic and environmental variables, diarrhea incidence can be predicted with high accuracy of 95.78 and 95.63% during training and testing phases, respectively. The model is useful for health policymakers in devising proactive intervention measures, including preparing healthcare systems and improving diarrhea prevention and control capabilities. It could also benefit future studies in predicting epidemics that are affected by similar variables.