LAUSR

COVID-19 is a contagious disease that has infected over half a billion people worldwide. Due to the rapid spread of the virus, countries are facing challenges to cope with the infection growth. In particular, healthcare organizations face difficulties efficiently provisioning medical staff, equipment, hospital beds, and quarantine centers. Machine and deep learning models have been used to predict infections, but the selection of the model is challenging for a data analyst. This paper proposes an automated Artificial Intelligence-enabled proactive preparedness real-time system that selects a learning model based on the temporal distribution of the evolution of infection. The proposed system integrates a novel methodology in determining the suitable learning model, producing an accurate forecasting algorithm with no human intervention. Numerical experiments and comparative analysis were carried out between our proposed and state-of-the-art approaches. The results show that the proposed system predicts infections with 72.1% less Mean Absolute Percentage Error (MAPE) and 65.2% lower Root Mean Square Error (RMSE) on average than state-of-the-art approaches.