Cloud-Integrated Machine Learning System for Ebola Virus Disease Prediction and Epidemic Intelligence in Smart Healthcare Systems

Abstract

Ebola Virus Disease (EVD) remains an important public health challenge, characterized by fast disease transmission and high mortality rates, as well as shortcomings in accepted outbreak monitoring systems. Existing prediction methods: (1) are not fully integrated into the cloud; and (2) do not analyze epidemiological, environmental and healthcare characteristics together in a single approach resulting in untrustworthy predictions. Hence, a framework which introduces hybridization of machine learning classification and deep-learning based temporal forecasting capable for Cloud-Integrated Machine Learning System resulting in Accurate Epidemic Intelligence with Early-Warning outbreak prediction is proposed. Experimental evaluation achieved more than a 98.92% accuracy, with precision and recall scores of over 98%, F1-score reaching almost at the upper range (and variations) and RMSE reducing prediction error successfully augments existing approaches in supporting timely effective intelligent decision making for healthcare systems.