A Predictive Model to Evaluate Stress in Students using Machine Learning

One of the prevalent tasks in the present scenario is to cope up with the mental stress resulting in a severe disease popularly known among all of us as depression. There exist numerous factors which are directly and indirectly responsible for increasing the statistics of people suffering from mental illness. Since earlier recognition of mental illness for example stress leading to depression could provide significant results. Therefore, this paper analyzes various features that could lead to severe mental illness to forecast the level of stress using machine learning techniques. The machine learning methods employed and compared in this research include Linear Regression (LR), Random Forest (RF), Support Vector Machine (SVM), Gradient Boosting Classifier (GBC), and K-Nearest Neighbor (KNN). In comparison, among the regression models, the best performance is given by Random Forest Regressor, which results in the lowest MSE value, 0.1292, as well as RMSE of 0.3594 and highest R² at 0.8065. These techniques are utilized and compared due to their capacity to examine correlations and hidden patterns in data. The findings reveal that the Random Forest model is a very effective predictor of stress levels, while Gradient Boosting provides a robust framework for classifying stress into various categories. This research paper presents and compares the performance of machine learning techniques to assess the stress level using a dataset based on standard measures.