Machine Learning for Forecasting Depression and Anxiety in University Students

The timely identification of mental health issues enables experts to more effectively provide treatment and enhance the well-being of patients. Mental health pertains to an individual’s emotional, mental, and interpersonal state, influencing their thoughts, emotions, and behaviors. It remains crucial across all life phases, spanning childhood, adolescence, and adulthood. Historically, categorizing mental health problems among college students demanded significant effort and time from psychologists. This research engaged five machine learning methods to classify such issues swiftly. The effectiveness of these methods was evaluated based on different standards. The five methods included logistic regression, KNN classification, decision tree classification, random forest, and stacking. A comparison and implementation of these methods revealed that stacking yielded the highest accuracy, predicting 82.20.