Introduction: The COVID-19 outbreak has nearly brought the globe to a standstill, and it has had both immediate and long-term effects on mental health university’s students. The current study aims to forecast changes in a few mental health indicators, including depression anxiety, social dysfunction, and loss of confidence among Palestinian medical students. Methods: The 300 students completed a General Health Questionnaire (GHQ) with a score of 15 or above. Afterward, the survey data was analyzed and sanitized. The survey data was examined, and a comparative prediction of the probabilistic changes of the mental health variables was carried out using common deep and machine learning techniques, such as deep Artificial Neural Network (DNN), Support Vector Machine (SVM), and Random Forest (RF). Results: The findings of these algorithms were reviewed using four commonly used statistical indicators to provide a better comparison between real and predicted data in terms of Coefficient of Determination (R2), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE). The DNN results were the best, with a coefficient of determination (R2) of 99% and the other error measures being 0.00002, 0.0046, and 0.0035 for MSE, RMSE, and MAE, respectively. The determination coefficient R2 for SVM and RF were 92.1% and 89.5%, respectively. Conclusion: This study highlights the importance of using machine learning tools for mental health prognosis.

Introduction: The COVID-19 outbreak has nearly brought the globe to a standstill, and it has had both immediate and long-term effects on mental health university’s students. The current study aims to forecast changes in a few mental health indicators, including depression anxiety, social dysfunction, and loss of confidence among Palestinian medical students. Methods: The 300 students completed a General Health Questionnaire (GHQ) with a score of 15 or above. Afterward, the survey data was analyzed and sanitized. The survey data was examined, and a comparative prediction of the probabilistic changes of the mental health variables was carried out using common deep and machine learning techniques, such as deep Artificial Neural Network (DNN), Support Vector Machine (SVM), and Random Forest (RF). Results: The findings of these algorithms were reviewed using four commonly used statistical indicators to provide a better comparison between real and predicted data in terms of Coefficient of Determination (R2), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE). The DNN results were the best, with a coefficient of determination (R2) of 99% and the other error measures being 0.00002, 0.0046, and 0.0035 for MSE, RMSE, and MAE, respectively. The determination coefficient R2 for SVM and RF were 92.1% and 89.5%, respectively. Conclusion: This study highlights the importance of using machine learning tools for mental health prognosis.