USING ARTIFICIAL NEURAL NETWORKS FOR PREDICTING CHICKENPOX CASES AT CHITUNGWIZA URBAN DISTRICT IN ZIMBABWE

Abstract

Despite the fact that chickenpox is rarely a fatal disease, it is still a real public health challenge; even in this 21st century. The current study used monthly time series data on chickenpox caseloads diagnosed and managed within Chitungwiza urban district in Chitungwiza, Zimbabwe, from Janaury 2012 to December 2019, to predict chickenpox cases over the period January 2020 to December 2021. We applied the well-known ANN (12, 12, 1) model. Residual analysis of this model indicated that the model is stable and thus suitable for forecasting chickenpox case volumes in Chitungwiza urban district over the out-of-sample period.