Psychological stress is encountered by a large number of people at varying levels; approximately 62% of individuals worldwide deal with mental stress daily. Students’ sensitivity to this type of phenomenon is significantly higher. The COVID19 pandemic has exacerbated this problem, with the pandemic causing a significant increase in mental distress worldwide. Given the magnitude of this problem, it is crucial to address and effectively reduce the causes of stress. Hence, the primary concern of the proposed study is to detect and quantify the level of stress among students. We applied a univariate feature selection method, specifically SelectKBest with a chi-squared test, and classified stress into three categories. In the dataset, we included 24 features collected from various studies and data sources on mental stress. For psychiatric disease screening, the Patient Health Questionnaire-9 (PHQ-9), a self-report questionnaire, was employed. For these 24 features, eight well-known conventional supervised machine learning algorithms were implemented, and the confusion matrices were analysed using key metrics to evaluate model accuracy. In the results section, with 87.3% accuracy, the highest prediction accuracy was attained by Support Vector Machines (SVM), highlighting that students are more prone to mild stress.