Artificial Intelligence-Enabled and Period-Aware Forecasting COVID-19 Spread

Coronavirus 2019 (COVID-19) has reached the stage of an international epidemic with a major socioeconomic negative impact Considering the weakness of the healthy structure and the limited availability of test kits, particularly in emerging countries, predicting the spread of COVID-19 is expected to help decision-makers to improve health management and contribute to alleviating the related risks In this article, we studied the effectiveness of machine learning techniques using Morocco as a case-study We studied the performance of six multi-step models derived from both Machine Learning and Deep Learning regards multiple scenarios by combining different time lags and three COVID-19 datasets(periods): confinement, deconfinement, and hybrid datasets The results prove the efficiency of Deep Learning models and identify the best combinations of these models and the time lags enabling good predictions of new cases The results also show that the prediction of the spread of COVID-19 is a context sensitive problem © 2021 International Information and Engineering Technology Association All rights reserved