Proposal of a Machine Learning Model for the Early Detection of Depression in University Students

Mental health problems such as depression in university students has increased with the passing of time, it is a topic little investigated by governmental institutions, but of great importance for the development of a country. The importance of early detection of mental illness would help to optimize the treatment process. Therefore, the overall objective of this research work is to detect depression early in college students by using a machine learning model, applying newer technologies and methodologies to develop a more robust and consistent model. The methodology used in this research consists of 5 phases: Dataset Extraction, Preprocessing, Feature Extraction, Model Implementation and Evaluation. Each phase was divided into activities and tasks to understand the problem-solving processes. In the results, the logistic regression (LR) technique obtained a superior performance with the metrics: precision of 0.79, sensitivity of 0.85, accuracy of 0.81 and f1 score of 0.81. In conclusion, machine learning techniques can detect depression in college students.