Abstract
The spread of COVID-19, namely SARS-CoV-2, has created a disastrous
situation around the world causing an unclear future. Machine Learning
(ML) and Deep Learning (DL) have a vital role in tracking the disease,
predicting the outgrowth of the epidemic, and outlining strategies and
policies to control its spread. Despite the inaccuracies of medical
forecasts, the numbers of COVID-19 cases forecasts provide us with
valuable information for recognizing the present and preparing for the
future. This study proposes a time series based deep learning model,
specifically the Long Short-Term Memory (LSTM) model. The model will
predict the active, confirmed, deaths and recovered cases for 7 days
ahead for Egypt and Saudi Arabia based on real-time data. The Egypt
prediction model achieves Mean Absolute Percentage Error (MAPE) of
3.26150, a Root Mean Square Error (RMSE) of 0.0144, a Mean Square Error
(MSE) of 0.0002, and a Mean Absolute Error (MAE) of 0.0092. While the
Saudi prediction model obtains a MAPE of 5.0553, a RMSE of 0.0170, a MSE
of 0.0002, and a MAE of 0.0150.