OPTIMIZED STRESS FEATURE EXTRACTION FROM COVID-19 DATA FOR MENTAL HEALTH APPLICATIONS

The COVID-19 pandemic has caused significant changes in people's lives, resulting in everyone suffering from mental health issues such as stress, financial pressure, depression, frustration, and anxiety. Identifying critical features associated with mental stress can help healthcare professionals to develop effective intervention strategies. This paper aims to design a machine learning-based decision support system (DSS) to assess the mental health status of an individual after COVID-19.The primary objective of this work is to give an in-depth statistical analysis and performance evaluation of machine learning for stress prediction, with the ultimate goal of mitigating the adverse effects of stress on mental health. A survey was carried out on around 1,200 individuals. The research finding shows that age and work area significantly impact mental health. The result analysis was presented for different machine learning approaches in which the Naive Bayes classifier and Logistic Regression achieved the highest accuracy of 99% whereas the Artificial Neural Network (ANN) and Support Vector Machine (SVM) achieved 71% accuracy. Random Forest shows a good performance of 98% and k-Nearest Neighbors (k-NN) shows 75% accuracy. The evaluation results indicate that logistic regression, naive Bayes, and random forest demonstrate superior performance. This research could lead to the development of stress prediction and prevention solutions based on a Decision Support System (DSS).