Stress is a physiological response of an organism to perceived threats or challenges. This study aims to identify students experiencing stress using Machine Learning algorithms. The proposed methodology comprises four parts: database acquisition, preprocessing, training, and evaluation. In this study, stress detection was performed using a classified database, implementing and training various Machine Learning algorithms, including a support vector classifier (SVC), Multilayer Perceptron (MLP), Logistic Regression (LR), K-Nearest Neighbors (KNN), Decision Trees (DT), Random Forest (RF), Gradient Boosting Machines (GBM), Adaptive Boosting (AdaBoost), Extreme Gradient Boosting (XGBoost), and light gradient boosting machine (LightGBM). The results demonstrate that the SVC model with parameters (C: 0.1; degree: 2; kernel: rbf), achieves an accuracy of 93%, a minimum accuracy of 95%, and recall of 90% for binary classification. Also noteworthy are the GBM and XGBoost algorithms, with parameters (learning_rate: 0.01; max_depth: 3; n_estimators: 50), both with an accuracy of 93% and precision of 93% and 94%, respectively. In summary, the SVC model is the best option for stress identification based on psychological, physiological, environmental, academic, and social factors.