摘要：This study aims to correlate weather variables with COVID-19 active cases by using Artificial neural network (ANN) method.Three types of Artificial Neural Network namely; Elman,NARX,and Feedforward networks,were designed and tested using MATLAB software.The COVID-19 active cases in Jordan data was obtained from the Ministry of Health in Jordan,while that of the weather data was obtained from Jordan Metrological Department.Both data were used in the development process of the models to approximate and estimate the actual performance of proposed models.The obtained results from training part were used to validate the ANN results.The performance of the three models was compared was decided based on the three statistics of metrology variables (R,RMSE,and MBE). Using the average daily temperature and wind speed as the input indicators to the network provided,it was found that Elman model exhibits the best performance and most accurate coefficient of correlation (R) and hence the most accurate correlation between weather variables and COVID-19 active cases in Jordan.while the predictive capability of Feedforward and NARX models were the least.
