Identifying Key Depressive Risk Factors Among Students Using Machine Learning Techniques.

Depression is rapidly emerging mental illness issue that still could not discover the major factor and reasons. School and college did not have any distinct solution regarding depression and number is growing very high over the year on depression among college students. Analysis in this paper is based on the dataset source from Kaggle, including information on sleep quality, study satisfaction, academic pressure, dietary habit, family mental health illness history, financial stress and suicidal thought. We determined that the most important factor related with depression risk are suicidal thought followed by academic pressure and financial stress. On the other hand, of those who had suicidal ideation, we estimate that 13,900 have depression and only 3,700 do not. Multiple classification models including Logistic Regression, Support Vector Machine (SVM), Random Forest, Naïve Bayes, Decision Tree, and K-Nearest Neighbors (KNN) were implemented using Python in Google Colab. Among them, Logistic Regression and SVM achieved the best results, with 84% accuracy and 88% recall, indicating strong predictive performance. The most significant factors contributing to depression were found to be suicidal thoughts, academic pressure, and financial stress.