Abstract

The way in which mental health affects university students has become an increasing concern, thus dictating the need for new strategies dealing with early identification and preventive methods. This research aims at evaluating the machine learning algorithms for predicting future mental well-being based on academic records and social media usage of students. A comprehensive dataset was examined using different techniques like Support Vector Machines (SVM), k-Nearest Neighbors (kNN), AdaBoost, Neural Networks, and Random Forest employing several algorithms. The outcomes indicated that most machine learning models were accurate with 96.7% accuracy noted across them all. Therefore, this demonstrates that machine learning methods effectively identify latent patterns and predict individuals’ personal mental changes among learners. These insights can be used to design appropriate interventions focusing on enhancing individual student’s psychological health within schools of higher education.