USING ARTIFICIAL NEURAL NETWORKS FOR PREDICTING SCABIES CASE VOLUMES AT CHITUNGWIZA URBAN DISTRICT IN ZIMBABWE

Abstract

Around the world, millions upon millions of people are affected by scabies, especially the disadvantaged populations living in crowded conditions in tropical areas. However, scabies is still a neglected tropical disease of the skin. Motivated by World Health Organization (WHO)’s recognition for the need for “a global strategy for scabies control”, the current study used monthly time series data on scabies caseloads recorded and managed within Chitungwiza urban district from Janaury 2012 to December 2019, to predict scabies cases over the period January 2020 to December 2021. We applied the famous ANN (12, 12, 1) model. Residual analysis of this popular model indicates that the model is very stable and thus suitable for forecasting scabies case volumes in Chitungwiza urban district over the out-of-sample period. The results of the study reveal that scabies cases will be on a relatively sharp downwards trajectory in Chitungwiza urban district over the out-of-sample period. The study, amongst other policy suggestions, encourages the government of Zimbabwe to always make sure that there is prompt administration of relevant drugs to infected patients as well as mite resistance monitoring programs in Chitungwiza urban district.