Predicting the COVID-19 spread, recoveries and mortalities rates in Saudi Arabia using ann

The worldwide pandemic of the COVID-19 has become the main national security issue for almost all nations The advancement of accurate prediction models provides insights into the spread of this infectious disease In fact, the high uncertainty and low size of data have caused some epidemiological models that show low accuracy for long-term prediction Although the related works include many attempts to deal with this issue, the robustness abilities of current models need to be enhanced In this paper, to achieve the main contribution, a prediction model using Artificial Neural Networks (ANNs) approach is developed based on the COVID-19 data from March 2, 2020 to August 5, 2020 to predict COVID-19 spread rate, recoveries rate, and mortalities rate in Saudi Arabia using Python programming language for the implementation stage and code has been developed to achieve the final results The evaluation in this paper has conducted through calculating the values for Correlation Coefficient (CC), Mean Absolute Error (MAE), and Mean Square Error (MSE) However, the results are promising by achieving low MAE with average value 0 05 and MSE with average value 0 02, and high Correlation Coefficient for all targets' rates with average value 0 97 Paper further recommends that real novelty in spread prediction can be realized through using other machine learning models with different types of COVID-19 data © 2005 - ongoing JATIT & LLS